JP2901675B2 - File backup acquisition synchronization method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION The present invention relates to a file backup acquisition synchronization method in a database sharing system in which a plurality of data processing devices access a common data resource, and more particularly to an improvement in the use efficiency of a computer. The present invention relates to a method for acquiring and synchronizing a file backup in a database sharing system suitable for the present invention.

[Conventional technology]

In a conventional online system, various methods have been proposed for efficiently performing database file recovery processing and improving computer use efficiency.

For example, database file recovery methods in a single online system with a single processor include:
A method for acquiring a backup of a database before starting online and while online is generally known.

Also, in systems that use a shared database by connecting multiple online systems via communication lines, in order to prepare for a file failure, a backup of the shared database is taken up before online starts, and a file journal is acquired. With the function, file recovery was performed using the merge journal created from all file journals.

In addition, a method is also conceivable in which an operator inputs a command to each online system and synchronizes backups by operating the system.

[Problems to be solved by the invention]

The above prior art does not discuss a method of synchronizing a plurality of data processing devices in a system sharing data resources. Therefore, when a plurality of online systems share a database file, there is a problem that a point for synchronizing the start of backup of the file cannot be set.

This is because when using the journal merge function, even if the identification of the file backup start point is added to the journal used as the merge input, recovery cannot be performed correctly for the following reasons unless synchronization is performed in each online system. It depends.

That is, if the file is updated before the identification is added by another online system after the identification is added by one online system, the file update information will be out of the target range of the journal merge function.

Further, in a method in which an operator inputs a command to each online system and synchronizes backups by operating the system, the load on the operator is large and an operation error may occur.

An object of the present invention is to improve such a problem and prepare for a failure of a database shared by a plurality of online systems. An object of the present invention is to provide a file backup acquisition synchronization method in a shared database system that can reduce the recovery time of a database by synchronizing the time points and targeting the journal after synchronization by a journal merge function.

[Means for solving the problem]

In order to achieve the above object, a file backup acquisition synchronization method of the present invention includes a plurality of data processing devices, a communication device interconnecting the data processing devices, and a file device storing data resources shared by the data processing devices. And a backup acquisition method for a database sharing system that accesses a shared data resource from a plurality of online systems. In the online acquisition system, a backup acquisition request command is input to any of the plurality of online systems while online, and the command is received. By suppressing the start of the new process, after the process in progress is completed, by notifying the online system other than the own system of backup preparation completion, and monitoring the synchronization status of the online system other than the own system, Synchronization of all online systems Since balanced point, releasing the activation suppression of new processing is characterized by performing the backup acquisition.

If the in-progress processing is not completed within a predetermined time, or if an online system other than the own system does not complete the backup preparation within a predetermined time, the backup acquisition request is canceled and the prohibition of the new processing is released. There is a feature in doing.

[Action]

In the present invention, in order to synchronize the backup acquisition point, the operator inputs a command (backup acquisition start request) to any one of the online systems, and this command is transmitted via a communication device interconnecting the data devices. Contact other online systems.

As a result, activation of a new process is suppressed, and all online systems are temporarily suspended.

Further, after waiting for the in-progress processing to be completed, the fact that synchronization has been achieved is notified to another online system. Thereafter, when all the online systems have been synchronized, an ID is added to the journal acquired thereafter, and the suppression of the start of the new process is released. When the backup acquisition is completed, a command for requesting the end of the backup acquisition is input to any of the online systems.

Therefore, a backup of the shared database is acquired from the start synchronization point to the end synchronization point, and in the event of a file failure, the file returned by the backup is subjected to recovery processing up to the time of failure by the file journal acquired after the start point. .

In other words, it is not necessary to perform backup of the shared database before starting online and recover the file from the beginning. It is only necessary to recover the file using the file journal since the start of online backup, improving the efficiency of computer use. At the same time, it is possible to build a 24-hour operation system.

〔Example〕

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

FIG. 2 is an explanatory diagram showing a file recovery procedure of the online system in one embodiment of the present invention.

2, reference numerals 21 and 22 denote data processing devices, 23 denotes an online system A operating on the data processing device 21, 24 denotes an online system B operating on the data processing device 22,
25 is a magnetic tape device for storing the file journal of the online system A, 26 is a magnetic tape device for storing the file journal of the online system B, 27 is a magnetic tape device for storing the merge journal, and 28 is a file device for storing the shared database , 29 are magnetic tape devices for storing backups.

In this embodiment, two online systems 23 and 24 operating on two data processing devices 21 and 22 are connected to a file device 28.
The following describes a case where the file journal is accessed by sharing the file journal.

First, each of the systems 21 and 22 acquires a backup of the file device 28 in the external device (magnetic tape device) 29 to prepare for a file failure.

Further, when a failure occurs in the file device 28, the file is returned from the backup of the external device 29, the file journals of the magnetic tape devices 25 and 26 are merged by the journal merge function, and a merge journal is created in the magnetic tape device 27.

After that, the file is recovered in the merge journal up to the point of failure.

Next, the synchronization processing procedure of the present embodiment will be described with reference to FIG.

FIG. 1 is a flowchart showing a synchronization processing procedure in one embodiment of the present invention.

In Figure 1, T ₁ is the time to monitor up tackle in process is completed, T ₂ is T ₁ - shows a (time tackle during processing required to finish). These are monitored by each online system in order to prevent a failure in a communication device interconnecting the data processing devices and leave the online system in a halt state.

In this embodiment, the operator makes a backup start request to the online system A23 shown in FIG. As a result, a backup start notification is sent to the online system B24 via a communication device (not shown) interconnecting the data processing devices 21 and 22.

When the backup start request is sent in this way, in the backup start process, the activation of a new process is suppressed (100).

Furthermore, wait for a certain period of time for
1) If the in-progress processing is completed within a certain period of time, the online system B24 is notified of backup preparation completion.

Thereafter, the completion of the preparation of the online system B24 is waited for a certain period of time (102), and the prohibition of the start of a new process is released with the time when the preparation of the online system is completed within the certain period of time as the synchronization start point (103).

If all of the in-process processes are not completed within a certain period of time, or if the online system B24 is not ready for backup, the backup acquisition request is canceled and the prohibition of the new process is released (104).

On the other hand, in the online system B24, when the backup start notification is received, the start of new processing is suppressed (20
0), and waits for the end of the in-progress processing for a predetermined time (201).

Further, if the in-progress processing is completed within a certain period of time, the backup preparation completion is notified to the online system A23, and the start of the new processing is released by setting the time when the online system becomes ready as the synchronization start point ( 202,203).

Therefore, access to the shared database is possible in the range X indicated by oblique lines, and access is disabled in the range Y.
In a range Z indicated by a mesh, file recovery is performed, and shared access is possible.

FIG. 3 is an explanatory diagram showing a comparison between a file recovery process according to an embodiment of the present invention and a conventional file recovery process.

In FIG. 3, (a) shows the conventional method, and (b) shows the method of the present embodiment. Reference numerals 31 and 32 denote file devices for storing databases.

In the conventional method (a), before starting the online system, a backup of the shared database stored in the file device 31 is obtained, and when a file failure occurs, the file journal up to the failure occurrence is stored in each online journal by journal merge processing. The system is merged, and the file is recovered from the beginning by the obtained merge journal.

On the other hand, in the method (b) of this embodiment, a backup of the shared database stored in the file device 32 is acquired after a synchronization point is set in response to a backup acquisition start request. From the backup acquisition start point using the merge journal obtained.

This eliminates the need to recover the file journal from the online start point to the backup acquisition start point, thereby greatly reducing the file recovery time.

〔The invention's effect〕

According to the present invention, in a system for accessing a shared data resource among a plurality of data processing apparatuses, a backup of the shared data resource can be acquired during the operation of the online system, so that the failure recovery time can be reduced.

In addition, since the time for which the backup was conventionally obtained can be used for other purposes, it is possible to improve the use efficiency of the computer and to construct a 24-hour operation system.

[Brief description of the drawings]

FIG. 1 is a flowchart showing a synchronization processing procedure in one embodiment of the present invention, FIG. 2 is an explanatory diagram showing a file recovery procedure of an online system in one embodiment of the present invention, and FIG. 3 is one embodiment of the present invention. FIG. 4 is an explanatory diagram showing a comparison between a file recovery process and a conventional file recovery process. 21, 22: Data processing unit, 23: Online system A, 24: Online system B, 25 to 27, 29: Magnetic tape unit, 28, 31, 3
2: File device, T ₁ : Time for monitoring until the in-progress processing is completed, T ₂ :, T ₁ − (time required until the in-progress processing is completed).

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Koji Mita 6-81 Onoe-cho, Naka-ku, Yokohama-shi, Kanagawa Prefecture Within Hitachi Software Engineering Co., Ltd. (56) References JP-A-59-177664 (JP, A) JP-A-62-2335 (JP, A) JP-A-62-86426 (JP, A) JP-A-63-136148 (JP, A) JP-A-63-254540 (JP, A) JP-A-63-259742 (JP, A) (58) Field surveyed (Int. Cl. ⁶ , DB name) G06F 12/00

Claims (2)

(57) [Claims] A plurality of data processing devices, a communication device for interconnecting the data processing devices, and a file device for storing data resources shared by the data processing devices; In the backup acquisition method for a database sharing system that accesses data resources, a backup acquisition request command is input to any of the plurality of online systems while online, and the online system receiving the command transmits the command to an online system other than its own system. Using the above communication device, notify the start of backup and suppress the start of new processing.After the processing in progress is completed, notify the online system other than the own system of the completion of backup preparation, and notify the online system other than the own system. Monitor system synchronization status A file backup acquisition synchronization method characterized in that, after all online systems have been synchronized, the suppression of the start of new processing is released and a backup is acquired. 2. If the in-progress processing is not completed within a predetermined time, or if an online system other than the own system does not complete the backup preparation within a predetermined time,
2. The file backup acquisition synchronization method according to claim 1, wherein the backup acquisition request is canceled and the prohibition of new processing is released.