JP2000082005A - Data processing system for inter-system data base sharing system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To sequentially input journals that respective data processors acquire at intervals of a certain time, to perform merging of the journals in a common- use data base, and to use them for subsequent processes without any delay. SOLUTION: Relating to data processors 1, 2 and 4 which share an inter- system data base, in communication information file devices 16 and 25 position information on obtained journals at intervals of a certain time is stored. Data processors which merge the journals that the data processors 1, 2 and 4 have obtained operate in parallel to the data processors 1, 2, and 4 to access the communication information file devices 16 and 25 at constant time intervals and receive the position information as the input ranges of the journals. When the journals are merged, the position information stored in the communication information file devices 16 and 25 is compared with the positions of the journals inputted through the merging process for the journals and journals up to the journal that the position information indicates are inputted to sequentially inputs the journals that the data processors have obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer system in which a plurality of data processing apparatuses access a shared data resource, and merges journals distributed and acquired on the plurality of data processing apparatuses to obtain subsequent data without delay. The present invention relates to a data processing method for providing to a processing process.

[0002]

2. Description of the Related Art A conventional data processing method for reflecting journal information in a subsequent data processing process in a computer system in which a plurality of data processing apparatuses access a shared data resource includes a journal acquired for each subsystem. In the file system for storing information, when the data set is full or the data set is replaced in any operation, the operator performs journal merge processing so that the operator can collect and merge the acquired journals of all the subsystems. Until the merged journal obtained in the journal merge process is provided to the subsequent data processing process, the process is performed in a batch process.

[0003]

According to the above prior art, journal merge processing is performed upon replacement of a data set storing journal information of all the subsystems,
This is a method to collectively provide journals after merging to subsequent data processing processes.Because journals are acquired in each of the above subsystems and journals are merged once for each medium via an operator, The problem that a delay occurs before the merged journal is provided for the subsequent data processing has not been considered.

According to the present invention, when a journal is acquired by each of the above subsystems, the journals are merged sequentially,
An object of the present invention is to perform efficient data processing by providing a merged journal to subsequent data processing. Another object of the present invention is to automate the operation of the journal merging process to prevent an operator's operation error, and to reduce the labor involved in operation.

[0005]

The above object has been achieved by the following means.

In each subsystem, a data processing device for storing position information in which a journal is stored in a file device at a fixed time interval or upon detection of occurrence of a specific event in a contact information file device and performing journal merge processing
(Hereinafter, referred to as journal merge) operates simultaneously with the subsystem, and receives, from the contact information file device, the journal position information stored in the file device by the subsystem as a range of journals that can be input at regular time intervals. In the journal merge, the position information stored in the contact information file device is compared with the position information of the last journal input by the journal merge, and if the position information stored in the contact information file device is not advanced, a fixed time interval is set. To access the contact information file device. If the location information stored in the contact information file device has advanced from the location of the last input journal, the journals acquired by the subsystem are input up to the journal indicated by the location information, the input journal is merged, and the Return to comparison. After that, the merged journal output by the merge processing is transformed into a format suitable for the subsequent data processing and delivered.

The effect of the processing of storing the journal position information in each subsystem on the subsystem is that the position information when the journal is stored in the file device is indispensable for the subsystem that has acquired the journal to manage it. Information, which is known information,
The access time required to store the location information in the file device is small because the size of the information is small, and the file device that stores the location information is occupied by each subsystem, so that the exclusive relationship with other subsystems is reduced. Need not be taken into account, and the interval at which the subsystem stores the position information in the file device can be extremely reduced by adjusting arbitrarily to the characteristics of the system.

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a block diagram of a typical computer system to which the present invention is applied. In FIG. 1, two data processing devices connected to each other via a communication device 3 are shown.
An example is shown in which two subsystems 13 and 23 operating on 1 and 2 share and access the file devices 14 and 24. Data processing devices connected to each other via the communication device 3 (hereinafter, referred to as a host or a host computer)
Includes one operating system (11, 21), one lock manager (12, 22) and one or more subsystems (13, 23). The shared data resource is stored in the shared file device (14, 24). Shared file device (14, 24)
Is connected to each host computer (1, 2). Each subsystem (13, 23) issues a lock request to the lock manager (12, 22), acquires a lock on the shared data resource, and then accesses the shared data resource via the operating system (11, 21). .

Each subsystem (13, 23) includes a journal file device (15, 25) and a contact information file device (16, 26). The subsystems (13, 23) perform processing in units called transactions, and record transaction history information in the journal file devices (15, 25). As a method in which a plurality of host computers access a shared data resource while performing exclusive control in block units, for example, as disclosed in JP-A-57-64850 and JP-A-57-64861, It is known that a plurality of host computers are connected to each other by a communication device to exchange lock information. Therefore, a description of exclusive control is omitted.

FIG. 2 shows the format of the contact information file stored in the contact information file device. Contact information file (5
1) is composed of a heading part (52) and a position information part (53). The heading part includes identification information (54) of each subsystem (13, 23),
It consists of operating state information (55). The position information section stores the position information of the journal stored in the journal file device (15, 25) by the subsystem (13, 23). The journal position information is composed of the identification information (56) of the device storing the journal and the journal block number (57). The subsystem (13, 23) stores the history information in the journal file device (1
The journal position information stored in (5, 25) is stored in the contact information file device (16, 26). The contact information file device (16, 26) records the identification information of the subsystem when the subsystem (13, 23) is activated, and the fact that the operation state of the subsystem (13, 23) is the initial state. Subsystem (13, 2
After the start of 3), the fact that the operation state is the execution state is recorded in the contact information file device (16, 26). Also, it stores the journal position information and the operation state information of the subsystems (13, 23), and updates them to the latest values at fixed time intervals defined for each subsystem (13, 23). Subsystem (13, 23)
At the end of the process, the fact that the operation state of the subsystem (13, 23) is the end state is recorded in the contact information file device (16, 26).

The data processing device 4 is a host computer
It is independent of (1, 2) and includes one operating system (41) and one journal merge (42). The journal merge (42) uses the contact information file device (16, 26) stored for each subsystem to store location information and subsystem (1
The journal of each subsystem (13, 23) stored in the journal file device (15, 25) is input using the operation status information of (3, 23) as a key, merge processing is performed, and the merged journal is merged into a journal output journal file. It is stored in the device (43). The procedure of the journal merge (42) process will be described with reference to FIGS. When the journal merge (42) is started, an initialization process is performed, and an input journal position is initialized (step 101). From the contact information file device (16, 26), the position information of the journal stored by all subsystems is input (step 102), and the input journal position is compared with the input journal position (step 103), and each subsystem (13, 23) It is determined whether the journal position information of the stored journal file device (15, 25) is advanced. If the position information obtained from the contact information file device does not advance from the input journal position, the next contact information file device (16, 26)
The process waits for an access to the server for a certain period of time (step 104), and then repeats the process of obtaining position information from the contact information file device. If the position information obtained from the contact information file device (16, 26) is ahead of the input journal position, the journal is merged (step 106). In the journal merge process (step 106), the journal of the subsystem indicated in the location information is merged with the journal file device (15, 25) until the end of the transaction is detected in order to merge the journal input in the transaction unit.
The data is input (step 204) and temporarily stored in a temporary storage area (step 205). When the end of the transaction is detected, the journal accumulated in the temporary area is merged (step 202), and the merged journal is stored in the merge output file device (43) (step 203). JP-A-63-201846 is publicly known as a method of merging journals acquired by a system sharing a database between systems, and a description of the merge processing method will be omitted.

Thereafter, the input journal position is updated.
(Step 107) Then, it is determined whether all the journals that can be input have been merged by comparing the input journal position with the position information (Step 105). If there are journals that can be entered, merge the journals (step 1
Repeat step 06). If there are no journals that can be entered, the operation status of all subsystems (13, 23) ends and the journal merge (42) processing is completed, or the subsystem operation obtained from the contact information file device (16, 26) A determination is made based on the state information (step 108), and if the operation states of all the subsystems (13, 23) are in the end state, the processing ends. If at least one of the subsystems (13, 23) is not in the end state, the journal position information stored by all the subsystems (13, 23) is input (Step 102), and the end of all the subsystems (13, 23) is completed. The process is repeated until is detected.

[0014]

According to the present invention, in a system for accessing a shared data resource among a plurality of data processing devices, journals acquired by respective subsystems can be provided to subsequent data processing sequentially while merging the journals. This is effective in that a simple database sharing system can be constructed. Further, according to the present invention, the operator can be automated without the intervention of the operator in the merge processing of the journals acquired by the respective subsystems, so that there is an effect that the operator's operation error can be prevented and the labor involved in the operation can be reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram of a typical computer system including the present invention.

FIG. 2 shows a format of a contact information file stored in a contact information file device.

FIG. 3 is a data configuration diagram of a contact information file.

FIG. 4 is a flowchart showing a journal merge processing procedure.

FIG. 5 is a flowchart showing a journal merge processing procedure;

[Explanation of symbols]

 1, 2, 4 ... data processing device 11, 21, 41 ... operating system 12, 22 ... lock manager 13, 23 ... subsystem 14, 24 ... shared file device 15, 25 ... journal file device 16, 26 ... contact information file Device 3 Communication device 42 Journal merge 43 Merge output journal file device

Claims (1)

[Claims] A plurality of data processing devices, a communication device interconnecting the data processing devices, a file device storing data resources shared by the data processing devices, and a system shared by the data processing devices. In a data processing system comprising a file device for storing the operation status of the system and a contact information file device for storing position information when the operation status of the system is stored in the file device, the data processing devices each access the shared data resource. And a plurality of online database access systems (hereinafter, referred to as subsystems) for performing the above-mentioned operations, and a data processing device for performing a merge process by comparing journals acquired by the respective subsystems with the corresponding subsystems. In the system, the above contact information file Merge processing is performed by sequentially inputting journals acquired by each subsystem using the identification name of each subsystem stored in the position information obtained from the storage device, the serial number of the medium, and the number of the block storing the journal as keys. A data processing system for an inter-system database sharing system characterized in that it is executed immediately after journal acquisition, and the merged journal is provided to subsequent data processing processes without delay.