JPH0581115A - Data processing method for distributed database managing system - Google Patents

Abstract

PURPOSE:To easily change management information, to reduce an error owing to change and to improve the reliability of a system at the time of moving the site of stored data by executing unitary management in the storage sites of data existing in the respective sites by means of the master site. CONSTITUTION:Management information of data in the respective sites WS 2-WSn is held as a data managing table 3b by separating it from information of the storage site and information concerning the storage site is executed unitary management as a site managing table 3a only in the master site WS1. The site managing table 3a is fetched from the master site WS1 at the time of starting the respective sites WS2-WSn and matching is executed with the data managing table 3b being in the self site at the time of retrieving data so as to obtain the storage site of data. Therefore, only the storage site name of the site managing table 3a is changed in order to operate the database system without a contradiction even if the storage site of data changes.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data processing method for a distributed database management system, and more particularly to a distributed database management system suitable for independently managing the data of each site and the sites independently. Data processing method.

[0002]

2. Description of the Related Art Conventionally, a data processing method in a distributed database management system of this type has been such that a database storage area for storing table data groups existing at each site is fixed to the site and a computer or database management system at each site is fixed. There is a method in which each site is uniquely identified by the site ID assigned to the site. This is, for example, as a nomenclature of tables distributed to sites, regarding management methods including site IDs, "Information Processing" Vol. 25, No. 5 (Showa 5).
9) pp. 443-450.

In the above conventional method, the table name for identifying the table data is expressed by a set of a creator name, a creator site name, an object name, and an object birth site name, and the object birth site, the storage site, and the past use are used. By storing the definition information of the table in the specified site, it is possible to determine the object storage site when the table is requested from any site. Also, under the control of the distributed database management system,
There is a function to move table data between sites.

An example of the data processing method of the conventional distributed database management system will be described below with reference to FIGS. 4 and 5. FIG. 4 is a configuration diagram of a general distributed database management system, and FIG. 5 is an explanatory diagram of a data structure used in the general distributed database management system. In the distributed database management systems Z1 to Zn, the workstations WS1 to WSn have databases DB1 to DBn and perform data processing via the transmission line L1. Further, in this example, the workstation management table 2, the own site ID 9, and the registration table 15 are present in the processors P1 to Pn, and the table data 10 is present in the databases DB1 to DBn. The site ID 9 is used to uniquely identify each workstation among all the workstations, and the registration table 15 is used to manage the table data group.

For example, when data is input from the terminal T3, the processor P existing in the workstation WS2.
2 communicates with the processors P1 to Pn existing in the other workstations WS1 to WSn, and determines which database the workstation having the data requested by the input data has.

This will be described based on the data structure of FIG. The information about the table is represented by a set of table name, creation date, object birth site name, and other necessary information. At each site, the contents are searched using the table name as a key. Further, when the table is generated, the table information is registered in the registration table 15 of the generated site. Further, the workstation names WS1, ..., WSn are used to identify the site.

Here, it is assumed that an inquiry request command for processing the table A and the table B is inputted from the terminal T3. The processor P2 of the workstation WS2 is
This inquiry request instruction is interpreted and it is recognized that the inquiry request instruction is processing table A and table B. Next, the registration table 15 is scanned, and the storage sites of the table A and the table B are compared with their own site ID 9 and checked. As a result, the storage site of the table A is determined to be the workstation WS2, that is, its own site, and therefore the process in the own site is performed.

On the other hand, since the storage site of the table B is found to be other than the workstation WS1, that is, its own site, the workstation management table 2 is scanned to obtain the communication line address of the workstation WS1, and the workstation WS1 is sent to the workstation WS1. A processing request for the table B is issued via the transmission line L1. If it does not exist in the registration table 15, the object birth site name is extracted from the inquiry request command. Since the table information should be registered in the created site registration table, a search request for an object storage site is made via the transmission line L1 based on the retrieved object birth site name.

In the distributed database management system, the data processing method as described above realizes the transparency of the table data group existing in each workstation so that the user is not aware of the workstation. There is. However, recent developments in hardware technology have made it possible to increase the capacity, size, and weight of a storage device used as a storage area of a database that stores table data, thus increasing the portability of the storage device. As ultra-small and lightweight storage devices such as floppy disks, memory cards, and IC cards become widespread in the future, it becomes unreasonable to consider these storage devices as fixed to the site. However, in the above distributed database management system, the databases DB1 to DB1
n is not fixed to the workstation,
There was a problem that normal operation could not be guaranteed.

[0010]

The prior art described above manages the distributed state of the table data based on only the site ID, and is an efficient method for changing the table data. Is not considered.

For example, the database DB1 is detached from the workstation WS1 and the workstation WS1 is removed.
When installed in No. 2, if the storage site WS1 of the table B in the registration table 15 is not changed to WS2, the registration table 15 becomes inconsistent and the table data included in the moved database DB1 is manipulated. Cannot be performed and the above operation cannot be performed normally. Therefore, when the database data is moved between sites, there is a problem that the registration tables of all sites of the database management system must be rewritten.

The present invention has been made to solve the above-mentioned problems of the prior art, and its purpose is to operate a distributed database management system without contradiction when moving data of a database between sites. The number of areas to be changed is reduced, the system reliability is prevented from lowering due to the error that tends to occur when the registration table is changed, and the site that stores the database data is saved to the users who use the workstations at each site. It is to realize the transparency of the position that does not make us conscious of.

[0013]

In order to achieve the above object, the structure of the invention relating to the data processing method of the distributed database management system of the present invention has a plurality of sites and a distributed system in which they are connected using a network. In a distributed database management system, a storage site having a master site and storing data of each site is centrally managed by a management number assigned to the storage site at the master site. Data processing method.

More specifically, in a distributed database management system in which a plurality of sites are combined with each other using a network, the present invention relates to a storage site that has a master site and stores management information of data at each site. It is a data processing method of a distributed database management system, characterized in that management is performed separately from information, and information about the storage site is centrally managed at a master site.

More specifically, in a distributed database management system in which a plurality of sites and those are combined using a network, a storage site having a master site and storing management information of data of each site The information is stored separately from the information regarding the storage site, the information regarding the storage site is centrally managed at the master site, and the storage information of the data is incorporated into each site to recognize the storage site of the data. It is a data processing method of a database management system.

More specifically, in a distributed database management system in which a plurality of sites and those are combined using a network, a master site is provided, and the management information of the data of each site is stored therein. It is managed separately from the information about the storage site, the master site holds a site management table having at least a management number and storage site items, and each site holds a data management table having at least a management number and a table name item, A data processing method for a distributed database management system characterized in that a site management table is taken into each site, and the site storing the data is recognized by matching the management numbers of the site management table and the data management table at the site. Is.

[0017]

The data management information at each site is held as a data management table separately from the information about the storage site, and the information about the storage site is centrally managed as the site management table only at the master site. Then, when starting each site, the site management table is fetched from the master site, and at the time of data search, it is matched with the data management table at the own site to obtain the data storage site. Therefore, even if the data storage site is changed, in order to operate the database system consistently, only the storage site name in the site management table needs to be changed, which is easy to modify.
And the amount of modification is small. Therefore, the risk of hardware errors and software errors that often occur when modifying the system management table is reduced, and the reliability of the system is improved.

Further, at each site, the contents of the site management table in the master site are automatically taken in at the time of startup, so that the workstation users at each site are aware of the data storage site. Needless to say, the transparency of the location where the data storage site of the distributed database management system is not recognized is achieved, the availability of the system is improved, and the system can be used more preferably.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment according to the present invention will be described below with reference to FIGS. 1 is a block diagram of a distributed database management system according to an embodiment of the present invention, FIG.
3 is an explanatory diagram showing a data structure of a distributed database management system according to an embodiment of the present invention, and FIG. 3 is a system configuration diagram of a distributed database management system according to an embodiment of the present invention.

First, an outline of the system configuration of the distributed database management system will be described with reference to FIG. Transmission line L
Workstations WS1 to WSn are connected to 1. In addition, the workstation WS1 has a terminal T
1 to T2 are connected and each terminal has a display device C1 to C
Holds 2. Similarly, terminals T3 to T4 are connected to the workstation WS2, and each terminal is a display device C3.
Through C4, the workstation WSn is connected to the terminal Tn, and the terminal Tn holds the display device Cn. Note that WS1 is referred to as a master workstation and is intended to contain special management information.

The workstation WS1 comprises workstations WS2 to WSn and terminals T1 to T.
2 can be communicated with. Similarly, the workstation WS2 can communicate with the workstations WS1 to WSn and the terminals T3 to T4, and the workstation WSn can communicate with the workstations WS1 to WS2 and the terminal Tn. For example, when the database related to the workstations WS2 to WSn is accessed from the terminal T1, the workstation WS1 transmits / receives data to / from the workstations WS2 to WSn via the transmission line L1 based on the data from the terminal T1. The result is then returned to the terminal T1 via the transmission line L1. The terminal T1 displays the above data on the display device C1. On the other hand, when the database related to the workstation WS1 is accessed from the terminal T1, the workstation WS1 accesses its own database based on the data from the terminal T1 without passing through the transmission line L1 and the terminal via the transmission line LT1. Returns a response to T1.
The terminal T1 displays the above data on the display device C1. The same is possible for workstations WS2 to WSn.

Next, how the distributed database management system according to the embodiment of the present invention is implemented in the distributed database management system shown in FIG. 3 will be described with reference to FIGS. .. For that purpose, first, FIG.
The data structure used in the data processing method of the distributed database management system of this embodiment will be described with reference to FIG. This is composed of a site management table 3a in which a management number and a storage site are paired, and a data management table 3b in which a management number, a table name and other information are paired. That is, the registration table 15 described in the related art is replaced with the site management table 3a.
And the data management table 3b, and each item is associated with the management number corresponding to each other. Now, set the site management table 3a to the master workstation WS1.
, The data management table 3b for each WS2, ..., WS
n workstations.

The operation of the workstation WS2 which is one site of the distributed database management system will be described below. First, workstation WS2
Site management table 3a for uniquely identifying each workstation in the file receiving unit 1 when the power is turned on
Via the transmission line L1 to the master workstation WS1
Receive files from.

For example, consider a case where the terminal T1 accesses a database related to the workstations WS3 to WSn. As an example, it is assumed that the table D in FIG. 2 is accessed. In this case, the question interpretation unit 5 of the workstation WS2 interprets the command received via the communication control unit 4 and recognizes that the data to be accessed accesses the table D. Then, the data processing unit 6 changes the table D from the data management table 3b.
The management number 004 of the item No. is searched, and then the site management table 3a fetched at the time of power-on is searched to recognize that the storage site of the item of the management number 004 is WS3. Thereafter, the data processing unit 6 compares it with the value WS2 of its own site ID 9 to identify that it is not a command for its own workstation. Therefore, the workstation management table 2 is scanned and the data transmission / reception unit 8 is used. Access the database of workstation WS3. Then, the data transmission / reception unit 8 transmits the accessed data to the terminal T1 via the communication control unit 4, and the terminal T1 displays the data on the display device C1.

On the contrary, consider a case where the terminal T1 accesses a database related to the workstation WS2. As an example, it is assumed that the table B in FIG. 2 is accessed. In this case, the question interpretation unit 5 of the workstation WS2 interprets the command received via the communication control unit 4 and recognizes that the data to be accessed accesses the table B. Then, the data processing unit 6
Searches the data management table 3b for the management number 002 of the item of the table B, then searches the site management table 3a fetched when the power is turned on, and recognizes that the storage site of the item of the management number 002 is WS2. To do. After that, the data processing unit 6 compares it with the value WS2 of its own site ID 9 and identifies that it is a command to its own workstation. Therefore, the table data 10 is accessed and the data is temporarily stored in the temporary storage unit 7. And the data is transmitted to the terminal T1 via the communication control unit 4. Then, the terminal T1 displays the above data on the display device C1.

Next, the effect of the present invention when the data of the database system is moved between the storage sites will be described using this embodiment. That is, in the prior art, the registration table 15 includes the storage site information. Therefore, if the data storage site is changed, it is necessary to change the registration table 15 at all the sites in order to operate the database system consistently. However, in this embodiment,
Information about the data is used as the data management table 3b,
Information regarding the storage site is separated and held as a site management table 3a. Therefore, for example, when changing the table A stored in the storage site WS1 to the storage site WS2, the corresponding management number 001 in the data management table 3b is searched and the item of management number 001 in the site management table 3a is searched. The storage site may be changed from WS1 to WS2. At each site, after restarting and taking in the changed site management table 3a, and when accessing the table A, look at the storage site of the item with the management number 001 in the site management table 3a, and find that it is in the storage site WS2. To know

As described above, conventionally, it was necessary to change the registration table 15 by the number of storage sites, but in the present embodiment, only one site management table 3a needs to be changed. Therefore, it is easy to modify the data storage site of the database system for changing the site.
Therefore, the risk of a hardware error or a software error that often occurs when modifying the system management table is reduced, and the reliability of the system is improved. Also, at each site, the site management table 3 in the master site WS1 at the time of startup
Since the contents of a are to be automatically imported, the workstation users at each site do not need to be aware of the data storage site, and the location where the data storage site of the distributed database management system is not aware of it. The transparency is achieved, the availability of the system is improved, and the system can be used more favorably.

Further, in case of a failure of the database system, only the site management table 3a need be restored for the information regarding the storage site. Therefore, there is also an advantage that recovery at the time of failure becomes easy.

[0029]

According to the present invention, when the database data is moved between sites, the distributed database management system is operated without any contradiction, so that the number of portions to be changed is reduced and the registration table is changed. It is possible to prevent the reliability of the system from being deteriorated due to an error that tends to occur, and to realize the transparency of the position where the user who uses the workstation of each site is not aware of the database data storage site.

[Brief description of drawings]

FIG. 1 is a block diagram of a distributed database management system according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram showing a data structure of a distributed database management system according to an embodiment of the present invention.

FIG. 3 is a system configuration diagram of a distributed database management system according to an embodiment of the present invention.

FIG. 4 is a configuration diagram of a general distributed database management system.

FIG. 5 is an explanatory diagram of a data structure used in a general distributed database management system.

[Explanation of symbols]

 L1 ... transmission line LT1-LTn ... transmission line C1-Cn ... display device D1-Dn ... database T1-Tn ... terminal P1-Pn ... processor WS1-WSn ... work Station Z1-Zn ... Distributed database management system 1 ... File receiving unit 2 ... Workstation management table 3a ... Site management table 3b ... Data management table 4 ... Communication control unit 5 ...・ Question interpretation section 6 ・ ・ ・ Data processing section 7 ・ ・ ・ Temporary storage section 8 ・ ・ ・ Data transmission / reception section 9 ・ ・ ・ Own site ID 10 ・ ・ ・ Table data 15 ・ ・ ・ Registration table

Claims (4)

[Claims] 1. In a distributed database management system in which a plurality of sites and those are connected using a network, a storage site having a master site and storing data of each site is set as a storage site at the master site. A data processing method for a distributed database management system, characterized in that the data is centrally managed by an attached management number. 2. A plurality of sites and a distributed database management system in which they are connected using a network, which has a master site, and management information of data of each site, and information about a storage site which stores the data. A data processing method for a distributed database management system, characterized in that separate management is performed and information about the storage site is centrally managed at a master site. 3. In a distributed database management system in which a plurality of sites and those are connected using a network, a master site is provided, and management information of data of each site is stored in the storage site for storing the data. A distributed database management system characterized by recognizing the storage site of the data by separately managing the information related to the storage site at the master site and incorporating the storage information of the data into each site. Data processing method. 4. In a distributed database management system in which a plurality of sites and those are connected using a network, a master site is provided, and management information of data at each site is stored as information about a storage site that stores the data. Separate management, a site management table that has at least a management number and storage site items at the master site, and a data management table that has at least a management number and table name items at each site, and site management at each site A data processing method for a distributed database management system, characterized by recognizing a data storage site by incorporating a table and checking the management numbers of the site management table and the data management table at the site.