JP2001290687A - Data-synchronization control system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data synchronization control system for automatically performing synchronizing processing after fault recovery and updating a data base in real time during synchronizing processing in a system connecting plural servers with incorporated data bases via a network. SOLUTION: In the system-connecting servers 10, 20 and 30 incorporated with data bases 13, 23 and 33 connected via a network 40, when data updating events occur in the servers 10, 20 and 30 before the recovery of the server 30 after fault occurrence, the data bases 13, 23 and 33 of the respective servers 10, 20 and 30 are updated, information on the differences of the data bases (such as device number or record number) is written on differential files 14, 24 and 34 and after fault recovery, while synchronizing processing of the data bases 13, 23 and 33 is performed, using the differential files 14, 24 and 34.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data synchronization control method, and more particularly, to a method for automatically updating a database when a new data update occurs during a synchronization process in a system for automatically synchronizing the database after recovery from a failure. The present invention relates to a data synchronization control method to be updated.

[0002]

2. Description of the Related Art A conventional database synchronization control method is disclosed, for example, in Japanese Unexamined Patent Application Publication No. 10-111825, entitled "Multiple Database Matching Update and Apparatus".
In this conventional technique, even if a new data update event occurs during the synchronization process, the data is not updated in real time (immediately).
Updated at the end of the synchronization process.

When a data update event is received while a failure occurs in another device, the data is temporarily stored in a difference file, and after the recovery from the failure, a database synchronization process is performed using this file. Also, when a data update event is received during the synchronization processing, a method of storing the data update event in the difference file is used. However, in the multitasking system, when a data update event is received immediately before the end of the synchronization process (at the moment when the communication status is returned from abnormal to normal because there is no event to be extracted from the difference file), communication is performed. The task determines that there is no data in the task that writes data to the difference file and the task that manages the difference file because the communication state is abnormal. Since the process of returning the communication state from abnormal to normal overlaps, the data is written to the difference file and the synchronization process ends without reflecting the content to another database, resulting in data mismatch.

[0004]

Therefore, the above-mentioned prior art is incapable of updating in real time even if a data update event occurs during synchronization processing at the time of recovery from a failure. The reason is that the data is updated after the synchronization process ends.

[0005]

SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide a system in which a plurality of databases are connected via a network and which require continuous operation and immediacy for 24 hours. It is an object of the present invention to provide a data synchronization control method that enables a database to be updated in real time when it occurs.

[0006]

SUMMARY OF THE INVENTION In a system in which a plurality of servers each containing a database are interconnected via a network or the like, the remaining servers are restored until one of the plurality of servers recovers from a failure. When a data update event has occurred, the database of a plurality of servers is updated, the difference information of the database is written to a difference file, and the database is synchronized using the difference file after recovery from the failure.

According to a preferred embodiment of the present invention, each server includes, in addition to the database and the difference file described above, a data update control unit for performing a database update process and a replication control unit for performing a database synchronization process. . The information of the above-mentioned difference file is a device number, a record number, and the like. Further, when a data update event occurs in one of the plurality of servers during the synchronization processing, the latest data of the updated record is set in one of the remaining servers, and a data registration notification is transmitted. Furthermore,
After updating the data, one server described above transmits a data registration completion notification to the remaining servers as a response corresponding to the data registration notification. After confirming that the data registration completion notification is normal, the remaining servers delete the corresponding data from the respective difference files.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The configuration and operation of a preferred embodiment of a data synchronization control system according to the present invention will be described below in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram showing the configuration of a preferred embodiment of a data synchronization control system according to the present invention. According to the data synchronization control scheme of this particular embodiment, it includes three servers 10, 20 and 30, each interconnected via a network 40. Each of the servers 10, 20, and 30 has a data update control unit 1
1, 21, 31 and the replication control units 12, 2
2, 32, and databases 13, 23, 33 and difference files 14, 24, 34.

In the data synchronization control system shown in FIG. 1, each server 10, server 20 and server 30 connected via a network 40 will be described.
The data update control unit 11 of the server 10 performs a database update process. The replication control unit 12 synchronizes the database 13 using the difference file 14. The database 13 manages data collectively. The difference file 14 stores difference information of the database 13. Further, the data update control unit 2 of the servers 20 and 30
1, 31, the replication control units 22, 32, the databases 23, 33, and the difference files 24, 34 have the same functions as the above-described data update control unit 11, replication control unit 12, database 13, and difference file 14, respectively.

Next, FIG. 2 shows servers 10, 20 and 3
1 and the difference files 14, 24, 34 included in FIG.
Parent table 1 included in databases 13 to 33 shown in FIG.
5 to 35 and child tables 16 to 36 are shown. The parent tables 15 to 35 and the child tables 16 to 36 included in the databases 13 to 33 of these servers 10 to 30 include:
It is assumed that referential integrity has been set.

Next, the operation of the servers 10 to 30 will be described with reference to FIGS.
This will be described with reference to the flowchart of FIG. First, the processing operation when each of the servers 10 to 30 is activated and operates normally will be described. It is assumed that a data update event (addition of data D11 and data D21) has occurred in server 10 (see step S1 in FIG. 4). The data update control unit 11 of the server 10 updates the content of its own database 13 (Step S2 in FIG. 4). After this database update, a data update notification is transmitted to the server 20 and the server 30 (see steps S3 and S4 in FIG. 4).
Data update control unit 21 of server 20 and server 30,
31 updates the contents of its own databases 23 and 33 (steps S5 and S6 in FIG. 4). Then, server 2
0 and 30 normally set and transmit a data update completion notification to the server 10 (step S in FIG. 4).
7 and S8). If the data update completion notification is normal, the server 10 terminates the process without performing anything. FIG.
The configuration of the databases 13 to 33 and the difference files 14 to 34 after the above-described processing is shown.

Next, the processing operation when the server 30 is not operating normally will be described. A data update event (addition of data D12 and data D22) occurs in server 10 (step S11 in FIG. 4). Data update control unit 1
1 updates the contents of its own database 13 (step S12 in FIG. 4). Then, the server 20 and the server 30
A data update notification is transmitted to the
13 and S14). Data update control unit 2 of server 20
1 updates the contents of its own database 23 (step S15 in FIG. 5). Thereafter, a data update completion notification is normally set and transmitted to the server 10 (step S1 in FIG. 5).
6). However, since the server 30 is in failure (step S9 in FIG. 4), the server sets an abnormality in the data update completion notification and transmits the notification, or terminates the process without doing anything (step S10 in FIG. 4).

The data update control unit 11 of the server 10 terminates the process without performing any processing since the data update completion notification received from the server 20 is normal. However, the server 30
Thereafter, if the data update completion notification is abnormal or a timeout occurs due to no response, the device number, record number, and update content are written to the difference file 14 and the process ends (step S17 in FIG. 5). Then, the server 20
Data update event (data D13 and data D2)
3) (step 18 in FIG. 5), and when a data update event (change data D23 to data D23α) occurs in the server 10 (step S25 in FIG. 5), the same processing as described above is performed (FIG. 5). Steps S19 to S26 of 5
And steps S27 to S31 in FIG. FIG. 3 shows the configurations of the databases 13 to 33 and the difference files 14 to 34 after the above-described processing is completed.

Next, the processing operation after the failure recovery of the server 30 (step S32 in FIG. 6) will be described. Server 3
The 0 replication control unit 32 transmits a replication start notification to the server 10 at the time of startup (step S33 in FIG. 6). The replication control unit 12 of the server 10 determines that the device number 3
Are extracted in ascending order (from the addition of data D12 and data D22), and a data update notification in which the latest data is set is transmitted to server 30 (step S34 in FIG. 6). Then, every time a data update completion notification is received from the server 30, the corresponding data is cleared from the difference file 14.

In the server 10, when a data update event occurs during the synchronization processing of the difference file 14, for example, the data D23α is newly added during the synchronization processing of the data D12.
Will be described (step S35 in FIG. 6) when a data update event occurs that changes the data D23β to the data D23β.
In this case, the database 13 of the server 10 is updated (Step S36 in FIG. 6). However, if the data update notification is transmitted in real time, the server 3
Since the data D13 does not exist in the table 35 of 0,
Referential integrity errors occur. Therefore, when a new data update event occurs during the transmission of the differential data, the data update control unit 11 of the server 10 notifies the server 20 of a data update notification (data D23α is transmitted to the data D23α).
23β) (step S37 in FIG. 6). Then, the database 23 of the server 20 is updated (Step S39 in FIG. 6), and a data update completion notification is issued (Step S40). The server 30 has a record number of 3
Latest data (data D13 and data D23β)
And sends a data registration notification (step S38 in FIG. 6). The data update control unit 31 of the server 30 updates the database 33 (Step S41 in FIG. 6). After this update, the server 30 becomes the server 10 and the server 20
A data registration completion notification is transmitted to the user (step S4 in FIG. 6).
2 and S43). The server 10 and the server 20 delete the corresponding data (the server 10 deletes the data D23α, and the server 20 deletes the data D13 and the data D23) from the difference files 14 and 24, respectively (steps S44 to S46 in FIG. 7).

Even if a data update event does not occur during the synchronization processing of the difference file 14, if the server 10 transmits the data D23α after the synchronization processing of the data D22, an error in reference consistency occurs. for that reason,
As described above, the sequence is to transmit a data registration notification to the server 30 instead of a data update notification. When the synchronization process between the server 10 and the server 30 is completed and all the difference files 14 are cleared, the replication control unit 12 of the server 10 transmits a replication end notification to the server 30 (Step S47 in FIG. 7).
The server 30 also transmits a replication start notification to the server 20 and performs the same processing (step S4 in FIG. 7).
8). When the server 30 receives the replication end notification from the server 20 (step S49 in FIG. 7), the synchronization of the database is completed (step S5 in FIG. 7).
0).

The configuration and operation of the preferred embodiment of the data synchronization control system according to the present invention have been described in detail.
However, it should be noted that such embodiments are merely examples of the present invention and do not limit the present invention in any way. It will be readily apparent to those skilled in the art that various modifications can be made without departing from the spirit of the invention. For example, the number of servers connected to the network can be any number equal to or greater than three.

[0019]

As apparent from the above description, the following practically significant effects can be obtained by the data synchronization control method of the present invention. First, even in a system in which three or more servers exist, the data synchronization processing can be automatically performed after a failure recovery. Second, even if a data update event occurs during data synchronization processing after recovery from a failure, the event can be updated in real time (immediately).

[Brief description of the drawings]

FIG. 1 is a block diagram showing a basic configuration of a preferred embodiment of a data synchronization control system according to the present invention.

FIG. 2 is a configuration example of a database and a difference file before a failure occurs in the data synchronization control method shown in FIG.

FIG. 3 is a configuration example of a database and a difference file after a failure occurs in the data synchronization control method shown in FIG. 1;

FIG. 4 is a part of a flowchart showing an operation of the data synchronization control method shown in FIG. 1;

FIG. 5 is a part of a flowchart following the flowchart shown in FIG. 4;

FIG. 6 is a part of a flowchart following the flowchart shown in FIG. 5;

FIG. 7 is a part of a flowchart following the flowchart shown in FIG. 6;

[Explanation of symbols]

 10, 20, 30 Server 40 Network 11, 21, 31 Data update control unit 12, 22, 32 Replication control unit 13, 23, 33 Database 14, 24, 34 Difference file 15, 16, 25, 26, 35, 36 Table

Claims (6)

[Claims] 1. A system in which a plurality of servers each containing a database are interconnected via a network or the like. When a data update event occurs in the remaining servers, the database of the plurality of servers is updated, the difference information of the database is written to a difference file, and the database is synchronized using the difference file after recovery from failure. A data synchronization control method characterized by performing a synchronization process. 2. The server according to claim 1, further comprising a data update control unit for performing update processing of the database and a replication control unit for performing synchronization processing of the database, in addition to the database and the difference file. The data synchronization control method according to claim 1. 3. The data synchronization control method according to claim 1, wherein the information of the difference file includes a device number, a record number, and the like. 4. The method according to claim 1, wherein one of said plurality of servers is performed during said synchronization processing.
4. The method according to claim 1, wherein when one data update event occurs, the latest data of the updated record is set in one of the remaining servers, and a data registration notification is transmitted. Data synchronization control method. 5. The one server, after updating the data,
5. A data registration completion notice is transmitted to the remaining servers as a response to the data registration notice.
Data synchronization control method described in 1. 6. The data synchronization according to claim 5, wherein the remaining servers delete the corresponding data from the respective difference files after confirming that the data registration completion notification is normal. control method.