JPH05181732A - File updating system in distributed processing - Google Patents

Abstract

PURPOSE:To effectively utilize this file updating system for large data capacity, variable data capacity, or processing requesting extreme shortage of shared time and occupied time by judging a restoring position based upon information stored in an updating frequency counter and then executing restoring processing. CONSTITUTION:A terminal equipment program 1 and a central processing unit(CPU) program 2 are mutually cooperated to update a center file 4. In the file updating system for distributed processing, information is stored in the updating frequency counter in each generation of updating of the file 4, and at the time of restoring data due to the generation of a fault, a restoring position is judged by the information stored in the counter 7 and its restoring processing is executed. Thereby the file updating system can be effectively utilized for large capacity like file updating for distributed transaction processing and processing requesting the extreme shortage of shared time and occupied time.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used in a computer system for distributed processing between terminals and central processing units,
In particular, the present invention relates to a file update method in distributed processing for updating files in distributed transaction processing.

[0002]

2. Description of the Related Art Conventionally, when processing is distributed between a terminal and a central processing unit to update a center file, as shown in FIG. 4, a terminal program 1 and a central processing unit program 2 are used.
The center file 4 is collectively updated at the end of a plurality of times of data transmission / reception 3 between and. Alternatively, as shown in FIG. 5, it is common to transfer all or part of the center file to the terminal program 1, process it on the terminal program 1 side, and retransfer to the central processing unit program 2.

These are characterized by a simple data recovery process when a failure occurs. That is, in the example shown in FIG. 4, the center file is collectively updated at the timing of the final data transmission / reception 3, and there is no need for data recovery in the event of a failure before that, and data recovery is possible only during collective update. do it. Further, in the example shown in FIG. 5, when a failure occurs, the data causing the failure may be discarded and reprocessed.

[0004]

However, in the file updating method in the conventional distributed processing, in the example shown in FIG. 4, the first and second transmission data must be accumulated in the memory. There is a drawback that effective processing cannot be performed when the capacity is large or when the data capacity is variable.

In the example shown in FIG. 5, since the data in the center file is exclusively used for a long time, it is difficult to use it in a transaction process which requires minimization of data sharing time and occupation time. There is a drawback.

The present invention solves such a problem, and when the data capacity is large, such as file update in distributed transaction processing, or when the data capacity is variable, and the sharing time and occupation of the data. An object of the present invention is to provide a file updating method in distributed processing that can be effectively used in processing that requires minimum time.

[0007]

In order to achieve this object, the present invention provides a file updating method in distributed processing in which a terminal program and a central processing unit program cooperate with each other to update a center file. Whenever a file update occurs, information is stored in the update count counter, and after that, when a failure occurs and data recovery is performed, it is possible to judge the recovery location from the information stored in the update count counter and perform recovery processing. It is a feature.

Further, the update counter is a terminal ID,
It is characterized in that a processing serial number and an update counter are continuously provided.

Further, the above-described processing and configuration are applied to file updating in distributed transaction processing of a computer system.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a file updating system in distributed processing according to the present invention will be described below with reference to the drawings.

FIG. 1 shows distributed transaction processing to which the embodiment is applied.

Here, as an example, data is transmitted and received three times between the programs of the terminal and the central processing unit, and the processing is completed by updating the center file respectively.

In FIG. 1, 1 is a terminal program, 2
Is a central processing unit program, and 3 is a data transmission / reception operation. Reference numeral 4 is a center file, 6 is an exclusive use flag, 7 is an update counter, 8 is a processing serial number counter, and 9 is a failure flag.

Next, the operation of this configuration will be described.

FIG. 2 shows the contents of the update counter 7. Further, FIG. 3 shows a recovery processing procedure executed by the central processing unit program 2.

In FIG. 1, when a transaction processing is requested from the terminal, the terminal program 1 assigns a processing serial number from the processing serial number counter 8 and sends it to the central processing unit program 2 together with the data.

In the central processing unit program 2, step a
In step 1, the exclusive use flag 6 is turned on to occupy the data used in the transaction, and the data is written to the center file 4. Also, terminal ID10, processing serial number 1
The data from the terminal program 1 is set in, and the update counter 7 is set to 1. As shown in FIG. 2, the update count counter 7 can determine the update count of the center file for each terminal and processing serial number.
After that, the processing up to this point is confirmed.

In response to this, the terminal program 1 transmits the second data to the central processing unit program 2. In the central processing unit program 2, after writing the data to the center file in step b as in the previous time, the update number counter 7 is set to 2 and the processing is confirmed.

Similarly, after writing the third data in the center file in step c, the update counter 7 is set to 2, the exclusive use flag is turned off, and the processing is confirmed. Transaction processing is performed in this manner.

Next, data recovery when a failure occurs and transaction processing ends abnormally will be described.

In FIG. 3, a request for restoration processing is judged from the terminal program 1 to the central processing unit program 2 (step (denoted by S in the figure) 100). In S100, if the request for the recovery process is No, the process ends. If the request for recovery processing is Yes, the process serial number counter 8 and the process serial number 1 are first compared (step 101). If the processing sequence number counter 8 has a larger No, it can be determined that the center file has not been updated at all, so the recovery processing ends (step 102).

Next, when it is assumed that the processing serial number counter 8 and the processing serial number 1 match, the update number counter 7
(Steps 103, 104, 10
5).

In step 103, the update counter 7 is set to 1
If Yes, the process of step a has been completed, so the process of step a is invalidated (step 10
4). After that, the update number counter 7 is set to 3 and the processing serial number is decreased to 1 (step 105).

When the result in step 103 is No, the update counter 7 is judged to be 2 (step 106). If the answer is 2 here, the processing in steps a and b has been completed, so the processing in steps a and b is invalidated (step 107). Then, the update number counter 7 is set to 3 and the processing serial number is decreased to 1 (step 10).
8). In the case of No in step 106, the update counter 7 is 3, and it can be determined that the update of the center file has been completed, so the recovery process is terminated (step 10).
9).

[0025]

As is apparent from the above description, in the file updating method in the distributed processing of the present invention, information is stored in the update counter every time a center file is updated, and thereafter, a failure occurs and data is lost. When performing restoration, the restoration location is determined based on the information stored in the update counter, and restoration processing is performed. Therefore, when the data volume is large, such as when updating files in distributed transaction processing, the data volume can be changed. In the case of, and the effect that it can be effectively used in the processing that requires the minimum sharing time and occupation time of data.

[Brief description of drawings]

FIG. 1 is a diagram showing program processing of a terminal unit and a central processing unit in an embodiment of a file updating method in distributed processing of the present invention.

FIG. 2 is a diagram showing the contents of an update counter in FIG.

3 is a flowchart showing a recovery processing procedure executed by the central processing unit program in FIG.

FIG. 4 is a diagram illustrating a program process of a terminal unit and a central processing unit, which is used for explaining a conventional example.

FIG. 5 is a diagram illustrating a program process of another terminal device and a central processing unit, which is used for explaining a conventional example.

[Explanation of symbols]

 1 Terminal Program 2 Central Processing Unit Program 3 Data Transmission / Reception Operation Path 4 Center File 6 Exclusive Use Flag 7 Update Counter 8 Processing Serial Number Counter 9 Failure Flag 10 Terminal ID

Claims (3)

[Claims] 1. In a file updating method in a distributed process in which a terminal program and a central processing unit program cooperate with each other to update a center file, information is stored in an update counter every time a center file is updated. After that, when a failure occurs and data recovery is performed, the recovery processing is performed by determining the recovery location based on the information stored in the update counter, and the file update method in distributed processing. 2. The file updating method in the distributed processing according to claim 1, wherein the update number counter is provided with a terminal ID, a processing serial number and an update number counter in succession. 3. A file updating method in distributed processing, wherein the processing and configuration according to claim 1 or 2 is applied to file updating in distributed transaction processing of a computer system.