JP2003203004A - Database management method for dual system and the dual system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technique capable of matching database of both lines without feeling the conditions of another line by updating the databases of both the lines without any communication between management systems. <P>SOLUTION: A master database 13 is provided for a management target 1. If data writing to the database 23 occurs in the management system 2 of a self line, the writing data is first written to the master database 13 and then written on the database 23 of the self line. The management target 1, after writing data to the master database 13, transmits a change notice including the writing data to a database 33 of another line to write the writing data to the database of another line. At the time of restoration time, the database is restored based on the master database 13 to match the databases 23, 33 of both the lines. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database management method in a dual management system.

[0002]

2. Description of the Related Art There are the following two conventional management system duplication methods. 1. The duplex configuration management system consists of an active system and a standby system, and normally uses only the active system. When the database is updated in the active system, the information is reflected in the standby system each time or periodically. When a failure occurs in the active system, the standby system is switched to the active system. 2. Both systems of the dual configuration management system are configured as the active system, and either system can be used. When a database update occurs in either system, the data is reflected in the other system in real time. That is, after writing to the database of the own system, it is transmitted to the management system of the other system to write to the database of the other system. When a failure occurs in one system, work is continued in the normal other system, and when the failed system recovers from the failure, a matching process is performed based on the database of the other system to match the databases of both systems. And resume business.

[0003]

In the above-mentioned conventional duplex configuration, since a processing procedure for switching the standby system to the active system is required when the active system fails, the system is in a stopped state from the failure occurrence to the completion of switching. There is a problem that becomes.

Further, in the dual configuration, the system does not become in a stopped state even when a single system fails due to the operation of both systems, but there are the following problems when the system is restored from the single system operation state. 1. When the failure of the other system is recognized by the disconnection of the communication with the other system, it is impossible to determine whether the other system is down or simply the communication is abnormal. Therefore, when a communication error occurs, both systems will recognize the other system as a failure.
In the case of dual configuration, when another system failure is recognized, each system independently updates the database of its own system without reflecting the information to the other system and continues to operate the system. When performing the matching process, the information of both systems may have been updated. In this case, it is difficult to completely reflect the database updates of both systems by the matching process.

From the above, the dual configuration is adopted to avoid the system stop at the time of one system failure, but the databases of both systems are updated without communication between the management systems. It has been desired to develop a database management method and a duplication system in a duplication system that can match the databases of both systems without being aware of the state of the system.

[0006]

A database management method in a duplicated system according to the present invention comprises a device to be managed and a duplicated management system connected to the device to be managed and each provided with a database. This is a database management method in a duplicated system in which a master database is provided in the managed device, and when writing to the database occurs in the management system of its own system, the write data is first written to the master database provided in the managed device. When writing, and then writing to the own system's database, the managed device sends a change notification containing the write data to the other system's management system to write to the other system's database, and the database consistency is required. Database based on the master database of the managed device. It is intended to match both system database by recovering.

In the present invention, a master database is provided in the managed device. Then, if data writing to the database occurs in your own management system,
The write data is first written in the master database, and then written in its own database. The managed device writes the data to the master database,
The write data is written to the database of the other system by transmitting a change notification including the write data to the database of the other system. That is, the databases of both systems are updated (matched) without transmitting / receiving data between the management systems. Then, at the time of restoration, the database is restored based on the master database, and the databases of both systems are matched.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a second embodiment of the present invention.
It is a block diagram which shows the structure of a duplication system. The duplication system 100 includes a management target 1 (only one device is shown in the drawing) which is an arbitrary independent device group, and management systems 2 and 3 having a device monitoring function. Managed object 1 is communication control unit 11, data processing unit 12, master database 1
It consists of 3. The management systems 2 and 3 are dual systems that operate independently of each other, and include communication control units 21 and 31, data processing units 22 and 32, and data bases, respectively.
It is composed of slots 23 and 33. The communication control unit 2 is provided between the management target 1 and the management system 2 and between the management target 1 and the management system 2.
1, 31 enable real-time communication. Unlike the conventional system, the management systems 2 and 3 are not connected to each other, and data is not transmitted / received between the management systems 2 and 3. 2 configured in this way
The duplication system 100 is specifically applied to, for example, a telephone exchange system that requires high reliability. In the case of this application example, the exchanges corresponding to the managed object 1 are managed by the management systems 2 and 3. Become.

FIG. 2 is a diagram showing a database write processing flow in the duplex system according to the embodiment of the present invention. First, the database write processing flow of the write processing of FIG. 2 will be described. When a database write occurs in the data processing unit 22 of the management system 2, the data processing unit 22 temporarily holds the write data inside (S1), and the communication control unit of the management target 1 via the communication control unit 21. The write data is sent to 11 as a write request message (S2). The data processing unit 12 of the management target 1 determines an event from the message received by the communication control unit 11. Here, it is determined to be a write request, and writing is performed in the master database 13 (S
At the same time, the write result is sent as a write response to the communication control unit 21 of the management system 2 via the communication control unit 11 (S4).

When the communication control unit 21 receives the write response, the management system 2 writes the internally stored write data in the database 23 (S5) and deletes the temporarily stored write data (S5). S
6).

On the other hand, the data processing unit 12 of the management target 1
After transmitting the write response to the management system 2, the management system 3 transmits the change notification including the write data to the communication control unit 31 of the management system 3 via the communication control unit 11 (S7).

When the data processing unit 32 of the management system 3 receives the change notification from the management target 1 via the communication control unit 31, it writes the write data included in the change notification into the database 33 (S8).

Similarly, the data processing unit 32 of the management system 3
If a database write occurs, the data processing unit 22 of the management system 2 transmits a message and writes the database in the opposite direction to the case where the database write occurs.

By the way, the duplex system 1 of the present invention adopts a dual configuration. Due to its characteristic, both management systems 2 and 3 simultaneously perform their own database write processing, and the management target 1 is processed. It is possible that write requests are sent simultaneously. In this case, in the duplex system 1, one of the write requests is accepted first, the step S6 of FIG. 2 is performed, and the other write request is accepted, and the steps S3 to S are performed.
The process of 6 is performed. Therefore, there is no difference in the databases 23 and 33 between the management systems 2 and 3, and the consistency can be maintained.

As described above, in the duplex system of this embodiment, when data writing occurs, the write data is transferred between the management systems 2 and 3 and the databases 23 and 33 of both systems are stored as in the conventional case. Instead of matching, the data transfer with other system is not performed at all, the write data is transmitted to the management target 1 and stored in the master database 13 first, and then the management system of the write generation source from the management target 1, By transmitting a write response and a change notification including write data to the management system of the system facing the write source, the databases 23 and 33 of both systems are updated. Therefore, there is no concept of "communication cutoff" with the management system of another system as in the past, and the downed system independently refers to the master database 13 of the management target 1 without being aware of the other system, It is possible to recover the database. Less than,
The flow of the matching process related to database restoration will be described with reference to FIG.

FIG. 3 is a diagram showing a database matching processing flow. For example, when the management system 2 goes down and the database 23 needs to be restored to match the database 33, the data processing unit 22 controls the communication control of the management target 1 via the communication control unit 21 of the management system 2. Part 11
A matching request message is transmitted to (S11). The data processing unit 12 of the management target 1 determines an event from the message received by the communication control unit 11, determines that it is a matching request here, reads all data from the master database 13 (S12), and reads the read data. As a matching response, the communication control unit 2 of the management system 2 is passed via the communication control unit 11.
1 (S13). The data processing unit 22 of the management system 2 confirms the data received by the communication control unit 21 (S
14), writing to the database 23 of the management system 2 (S15). As a result, the database 23 is restored based on the master database 13, and the database 2
2, 23 are matched to the latest data.

Further, the database 33 of the management system 3
When the data is matched with the database 22, the data processing unit 22 of the management system 2 causes the communication control unit 2 of the management system 2 to
The matching request message is transmitted to the communication control unit 11 of the management target 1 via 1. Data processing unit 12 of management target 1
Judges an event from the message received by the communication control unit 11, judges that it is a matching request here, reads all data from the master database 13, and sends the read data via the communication control unit 11 to the management system 3. To the communication control unit 31. The data processing unit 32 of the management system 3 confirms the data received by the communication control unit 31 of the management system 3 and writes it in the database 33 of the management system 3.

In order to simplify the above description, FIG. 1 illustrates an example in which the managed object 1 and the management systems 2 and 3 each have one database, but in practice, A plurality of databases (the number of databases provided in each of the management target 1 and the management systems 2 and 3 is the same), and in step S14,
Confirm that the data of all databases can be received. If it has not been received, the process returns to step S11, and the matching request is transmitted to the management target 1 for the database that could not be received. In other words, when performing database matching, all databases are adapted at once.

Next, the flow of the database reading process will be described with reference to the drawings. FIG. 4 is a diagram showing a database read processing flow. When the data processing unit 22 of the management system 2 needs to read the database,
The data processing unit 22 of the management system 2 uses the database 23.
Is read (S21). Similarly, when the data processing unit 32 of the management system 3 needs to read the database, the data processing unit 32 of the management system 3 reads the database 33.

According to the present embodiment, when the master database 13 is provided in the management target 1 and the writing processing occurs in the databases 23 and 33 of the management systems 2 and 3, the master database 13 is written first, After that, the following effects are expected by performing writing to the databases 23 and 33 of the management systems 2 and 3 based on data transmission (write response, change notification) from the management target 1 to each management system 2 and 3. it can.

By using the master database 13 of the management target 1 as a master database, the management systems 2 and 3
When starting up and when the management system 2 or 3 recovers from a special case such as single system operation, the management system performing the recovery acquires data from the master database 13 It is possible to match the database of. That is, it is possible to bring the databases of both systems into agreement with the latest information by restoring the databases of its own system based on the master database 13 without being aware of the states of the other systems.

In addition, when the management system 1 and the management system 1 simultaneously write their own databases and write requests are sent to the management target 1 at the same time,
The management target 1 receives one of the write requests first, performs the processing up to step S6 in FIG. 2 and then receives the other write request, and performs the processing from step S3 to step S6. Database 2
There is no difference between 3 and 33, and the consistency can be maintained.

When the database is updated, the management target 1
The master database 13 is updated, and the database of the other management system is updated by the change notification from the management target 1. Communication is not required, and the management systems 2 and 3 can continue to manage the management target 1 independently of each other regardless of the states of the management systems of other systems. Further, since communication between the management systems 2 and 3 at the time of updating the database becomes unnecessary, it is possible to reduce communication cost.

Conventionally, there is a technique in which a database is not provided in a management system, but a database serving as a master is provided only in a management target. In this conventional technique, when a database is read, the master provided in the management target is used. The response is bad because it needs to be read from the database. On the other hand, in the present invention, since the database is also provided in the management system, the database of the management system can be used during reading, and the reading time can be shortened. Further, as compared with the existing dual configuration system described in the section of the prior art, the present invention can achieve the above effect while maintaining the database read speed which is the same as the database read speed in the existing system. It can be said that this is a useful technique.

In the above description, the case where the management system is duplicated has been described as an example, but since the management system can access the database without being aware of the operating status of the management systems of other systems, the management target 1 By increasing the number of connectable management systems on the side, it is possible to multiplex more than double.

[0026]

As described in detail above, according to the present invention, a master database is provided in a device to be managed, and when a database write occurs, the write data is first written in the master database and then each management is performed. Since the databases of both systems are updated without transmitting / receiving data between the systems, the database of both systems can be matched without being aware of the status of other systems by restoring the database based on the master database at the time of restoration. Is possible.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a duplex system according to an embodiment of the present invention.

FIG. 2 is a diagram showing a database write processing flow.

FIG. 3 is a diagram showing a database matching processing flow.

FIG. 4 is a diagram showing a database read processing flow.

[Explanation of symbols]

1 management target A few management systems 12,32 Data processing unit 13 Master database 23,33 database

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Mitsuru Kato             1-7-12 Toranomon, Minato-ku, Tokyo Oki Communication             System Co., Ltd. (72) Inventor Kiyoshi Okada             1-7-12 Toranomon, Minato-ku, Tokyo Oki Communication             System Co., Ltd. (72) Inventor Masato Kojima             1-7-12 Toranomon, Minato-ku, Tokyo Oki Communication             System Co., Ltd. (72) Inventor Koji Yokochi             1-7-12 Toranomon, Minato-ku, Tokyo Oki Communication             System Co., Ltd. F term (reference) 5B034 AA01 BB01 DD05 DD06                 5B082 DE03

Claims (2)

[Claims] 1. A database management method in a duplicated system comprising a managed device and a duplicated management system connected to the managed device and provided with a database, respectively. If a master database is installed in the management system and writing to the database occurs in the management system of the own system, write data is first written to the master database provided in the managed device, and then written to the database of the own system, and the management is performed. When the target device sends a change notification containing write data to the management system of other system and writes it in the database of other system, and when the database consistency is required, it is based on the master database of the managed device. Restoring the database ensures that both databases are consistent. Database management method in a duplex system according to claim. 2. A duplicated system comprising a managed device and a duplicated management system connected to the managed device and each having a database, wherein the managed device is a management system A master database for storing the received write data, a unit for writing the write data received from the management system to the master database, and a write result of writing the write data in the master database to the management system of the write data transmission source. Along with, a means for transmitting a change notification including the write data to a management system facing the transmission source, and means for reading out and returning all data in the master database based on a matching request from the management system, The management system can write to its own database Means for temporarily holding the write data when it occurs, a means for transmitting the temporarily held write data to the managed device, and a means for storing the temporarily held write data based on the write result received from the managed device. A means for writing to the system's database and a means for writing the write data included in the change notification to the own system's database when the change notification is received from the managed device, and the own system's database must be matched with the databases of other systems If there is, a means for transmitting a matching request to the managed device, restoring the database of the own system based on the data from the managed device based on the matching request, and matching with the database of the other system are provided. Duplex system characterized by.