JPH05233342A - Job restart control system - Google Patents

Abstract

PURPOSE:To dispense with the rerun of an interrupted job from the first, thereby, to delete the consumption of wasteful time. CONSTITUTION:A restart file 2 which stores restart information in which a job name, a file name, and a restart address are recorded is provided, and when the update processing of data in a data base(DB) 3 is performed, processing to write a new address on the restart address of the restart information in accordance with its own job 1 in the restart file 2 is performed at an appropriate interval, and when a certain fault occurs in a computer, and a data update operation is interrupted, and the fault is recovered, then, the job 1 is restarted, the update operation is immediately performed from the address recorded on the restart address of the restart information in accordance with its own job 1 by referring to the restart file 2.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a job restart control system for recovering a fault and restarting the job when a fault occurs during execution of a job and the job is interrupted.

[0002]

2. Description of the Related Art In a computer system, when a failure occurs during the execution of a database (DB) job, the execution of the job is interrupted, the failure is recovered, and the job is restarted. The start control method is based on the database (D
Restore the entire file in B) to the state before executing the job and re-execute the job from the beginning, or leave the database (DB) file in the state in which it was interrupted, and then restart the interrupted job from the beginning. A method is adopted that re-executes and skips the updated data before the interruption and updates the unupdated part of the file.

[0003]

As described above, according to the conventional job restart control method of the database (DB) system in the computer system, when the failure is recovered and the job is restarted, the interrupted job is first processed. Since it has to be re-executed, it has a drawback that it takes a considerable time to reach the interruption point and wastes time.

[0004]

According to the job restart control method of the present invention, a restart file for storing restart information for recording a job name, a file name and a restart address is provided to update data in a database. When executing the job, the restart address is periodically written in the restart information corresponding to the own job of the restart file, and some failure occurs in the computer system to interrupt the data update operation,
When the fault is recovered and the job is started again,
It includes restarting the update operation immediately from the restart address of the restart information corresponding to the own job by referring to the restart file.

[0005]

Embodiments of the present invention will now be described with reference to the drawings.

FIG. 1 is a schematic diagram showing an operating state of an embodiment of the present invention, and FIG. 2 is a format diagram showing an example of contents of restart information in the restart file of FIG.

In FIG. 1, a job 1 is a job for updating the data in the database (DB) 3, and the restart information is written in the restart file 2 during the update process. It has a built-in process to write periodically.

As shown in FIG. 2, the restart information 4 is
The job name 41, the file name 42, and the restart address 43 are included. The job name 41 and the file name 42 manage the restart information for which file of which job.
3 controls where to start when restarting. The restart information of a plurality of jobs can be stored in the restart file 2.

The restart information 4 is initialized in advance in the restart file 2. That is, job name 4
1 indicates the name of the job that is the target of restart management.
The file name 42 is the name of the file updated by the job, and the restart address 43 is the file name 42.
Addresses for which the job recorded in the job name 41 first updates the data in the file recorded in 1 are set as their initial values.

In a normal process, the job 1 reads the restart information 4 corresponding to its own job by referring to the restart file 2 when starting the operation, and the restart address 43 having the initial value set therein. The data update is started for the database (DB) 3 according to the information.

The job 1 is restarted corresponding to its own job in the restart file 2 at appropriate intervals (for example, for each fixed amount of data) while updating the data of the database (DB) 3. The process of writing a new address in the restart address 43 of the information 4 is repeated.

Job 1 is a database (D
B) When any failure occurs in the computer system during the data updating operation of 3, the job 1 interrupts the data updating operation at that time. When the failure is recovered and the job 1 is restarted, the job 1 refers to the restart file 2 to read the restart information 4 corresponding to its own job, and is recorded in the file name 42 and the restart address 43. The restart address for the corresponding file is known from the contents described in the table, and the update operation is restarted immediately from that address.

[0013]

As described above, according to the job restart control method of the present invention, a restart file for storing restart information for recording a job name, a file name and a restart address is provided, and a database (D
At the time of performing the data update process of B), at a proper interval, a process of writing a new address into the restart address of the restart information corresponding to the own job in the restart file is performed to cause some trouble in the computer system. When the error occurs, the data update operation is interrupted, the failure is recovered, and the job is restarted, the restart file is referenced and the address recorded in the restart address of the restart information corresponding to the own job is changed. By directly restarting the update operation, it is not necessary to re-execute the interrupted job from the beginning, and thus it is possible to eliminate wasteful time consumption.

[Brief description of drawings]

FIG. 1 is a schematic diagram showing an operating state of an embodiment of the present invention.

FIG. 2 is a format diagram showing an example of contents of restart information in the restart file of FIG.

[Explanation of symbols]

 1 Job 2 Restart File 3 Database (DB) 4 Restart Information 41 Job Name 42 File Name 43 Restart Address

Claims (2)

[Claims] 1. When a job for updating a database is executed, a restart address is regularly written in the restart information corresponding to the own job in the restart file, and some failure occurs in the computer system. When the data update operation is interrupted, the failure is recovered and the job is started again, the update operation is immediately restarted from the restart address of the restart information corresponding to the own job by referring to the restart file. A job restart control method including: 2. A restart file for storing restart information for recording a job name, a file name, and a restart address is provided, and when a job for updating data in a database is executed, the restart file itself is stored. The restart address is written in the restart information corresponding to the job periodically, and some kind of failure occurs in the computer system to interrupt the data update operation,
When the fault is recovered and the job is started again,
A job restart control method comprising: restarting the update operation immediately from the restart address of the restart information corresponding to its own job with reference to the restart file.