JPS6337438A - Site control system for distributed database control system - Google Patents

Abstract

PURPOSE:To operate table data even if a storage device including a database storage area storing a table data group moves between sites by providing a site controller. CONSTITUTION:The site controller is comprised by combining a site ID and a database ID, and controls the table data groups in the sites. When a command to register data in the database storage area is inputted from a terminal 4-1, a registration table controlling a data table and a self database ID are read in from a secondary storage medium 37 in a database 5-1 through a site control program 33. They are set to a registration table storage part 35, and simultaneously the database ID is registered in a database control table. The registered database ID and the site ID are reported to other sites. They set both the IDs in their own database ID control tables. Even if the medium 37 is removed from a secondary storage medium controller 36 to install in other devices, the table data can be operated.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to distributed management of data in a distributed database management system, and in particular to a site management method suitable for distributed management of database storage areas independently of sites. Regarding.

[Conventional technology]

Conventionally, the site management method for this type of distributed system has been to fix the database storage area for storing the table data group existing at each site at each site, and to use a unique identifier assigned to each site's computer or database management system. This is done based on possible site IDs.

For example, regarding the management system including site IDe as a nomenclature for tables distributed across sites, "Information Processing" 2nd Edition
Discussed in Volume 5, No. 5 (1981), pages 443 to 450. In the conventional device described above, a table name for identifying table data is expressed as a set of a creator name, a creator site name, an object name, and an object birth site name. This shows that by retaining table definition information, the object storage site can be determined when a request for the table is made from any site. It also provides the ability to move table data between sites under the control of a distributed database management system.

However, recent developments in hardware technology have made it possible to increase the capacity, size, and weight of storage devices used as storage areas for databases that store table data, thereby increasing the portability of storage devices.

In the future, ultra-compact and lightweight storage devices such as floppy disks, memory cards, and IC cards will become widespread.

It becomes difficult to make these storage devices fixed to the site.

[Problem that the invention seeks to solve]

The above conventional technology manages the distributed state of table data based only on the site, and the storage device containing the database storage area for storing the table data group is removed from the site and set at the other site. There is no consideration given to the fact that the storage device itself is moved between sites, such as when the storage device is moved between sites, and there is a problem that the table data group contained in the storage device cannot be searched when the storage device is moved between sites. there were.

An object of the present invention is to realize a site management method for a distributed database management system that allows table data to be manipulated even when a storage device including a database storage area for storing table data n is moved between sites. It is in.

[Means for solving problems]

The above purpose is to manage table data existing at each site in a distributed database management system that handles table data groups scattered on a system configured using multiple sites and a communication line network that connects them. In order to uniquely identify each site, the site management method for
Site management that controls the table data group of each site by combining the site ID (site identifier), the database IDC (database identifier), and the site ID and database IDt for uniquely identifying the database storage area that stores the table data group existing at the site. This is accomplished by providing and configuring a device.

[Effect]

In the above invention, the site management device monitors the activity status of each site constituting the distributed database management system at site I.
The DK manages the activity status of the database storage area that stores table data # installed at the site using the database ID. When a site starts its activities, a message from the site is sent to the site management device to request activation of the site. Thereby, the site management device registers and activates the site II). Database storage area on the site? When the containing storage device is installed, the site sends the database ID to the site management device,
Request activation of the database. Thereby,
The site management device registers and activates the database in association with the site ID. On the other hand, the table naming method for grouping table data uses database IDs without using site IDs.

When an inquiry request for table data occurs in the above state, the site management device can determine the site where the database ID is activated based on the database ID included in the inquiry. will not malfunction even if you move between sites.

〔Example〕

An embodiment of the present invention will be described below with reference to FIGS. 1 to 4.

FIG. 2 is an example of a distributed database system using a distributed database management system. In the figure, 1 is a communication network, 2-1 to 2-3 are processors, and 3-1 to 3-
3 is a distributed database management system, 4-1 or 4-
3 represents a terminal, and 5-1 to 5-3 represent databases storing table data groups.

In the molecule, tf-tapace system, each processor 2-1 to 2-3 is connected to a database management system 3.
-1 to 3-3, and performs data processing of each database 5-1 to 5-3 via one communication network 1. for example,
A database management system 3- existing in the processor 2-1, into which the transaction cushion is input from the terminal 4-1.
1 is the processor 2-2 or 2- of the cytotra pond where the table data requested by the transaction cushion exists.
Database management system 3-2 or 3 existing in 3
-3 and make a decision.

An example of a conventional site management system will be explained with reference to FIG. In the figure, 10 is a site management table, and 11 is a site I for uniquely identifying each site among all sites.
D, 12 is a registration table for managing a group of table data, and 13 is table data. In this example, the site management table IO and own site IDI 1 are stored in the registration table 12 in the process server 2-1 or the database management system 3-1.
and table data 13 exist in database 5-1. Site B or site OK also has similar table configuration information.

Assume that an inquiry request command for processing tables A and B is input from the terminal 4-1.

It is assumed that the table name is expressed as a set of table creator name, creator site name, object name, and object birth site name.

The database management system 3-1 interprets the above command and recognizes that the inquiry request command targets tables A and B. Next, registration table 12
is scanned and compared with the stored site controller's own site IDI 1 in tables A and B. As a result, it is determined that the storage site for table A is site A, that is, the own site, so processing is performed within the own site. On the other hand, it turns out that the storage site for table B is site B, that is, other than the own site, so the site management table 10 is scanned to obtain the communication line address of site B, and the storage site for table B is sent to site B via communication network 1. Make a processing request.

If the object does not exist in the registration table 12, a search request is made for the object storage site via the communication network 1 based on the internal creator site name or the object creator site name in the table name.

In the distributed database management system, by using the site management method described above, the table data group existing at each site is made transparent to the user without making the user aware of the site.

On the other hand, in the above-described distributed database system, normal operation of the databases 5-1 to 5-3 cannot be guaranteed unless the site K is fixed. According to the above example, if the database 5-1 is removed from site A and installed at site B, an inconsistency will occur in the registration table 12, making it impossible to manipulate the table data included in the moved database 5-1. In order to be able to execute the above operations normally, it is necessary to refer to the tables included in the database 5-1 to be moved and the databases 5-2 to 5-3 of other sites that referenced the database 5-1 of site A. , sends a notification from the site user that they will be moving to site H,
Other sites have to change the contents of the related registration table according to the notification, resulting in an inefficient system. As the number of tables or sites increases, the amount of communication between sites increases synergistically, increasing message transfer costs.

Therefore, in the case of the present invention, (1) a database ID is provided independently from the site, (1i) the database ID is added to each database storage area that stores the table data group and the registration table, and the GiD database ID and database storage The areas are associated one-to-one and stored in the storage device in this unit. On the other hand, the Rei Registration Table attempts to solve the conventional problems by adding a storage database storage area in place of the storage site ID and a database ID table separate from the M site management table.

FIG. 4 shows an embodiment of the site management method according to the present invention. In the figure, site management table 10, 11 from the own site
, table data 13 corresponds to FIG. 2, but the others are added according to the present invention, 20 is a database ID management table, and 21 is a registration table for managing table data.

Reference numeral 22 indicates a database ID (database identifier) added to the database storage area in which the registration table 21 and table data 13 are stored.

The registration table 21 in the system of the present invention has been changed to hold the storage database ID instead of the storage site ID item of the registration table 12 in the conventional system. Furthermore, like the site management table 10, the database ID management table 20 in the method of the present invention is held in the processor 2-1 or the database management system 3-1 of site A, and is a database ID management table installed at each site.
is managed. Further, the own database ID 22 is held in the same storage device that stores the database storage area that stores the registration table 21 and the table data 13,
It is determined so that all database IDs can be uniquely identified. There may be cases in which the database storage area in the present invention increases and cannot fit into one storage device, but in this case, it can be easily handled by adding the volume S of the storage device to the own database ID 22. .

FIG. 1 shows an example of processing using the site management system of the present invention, using site A in FIG. 2 as an example. To explain the new elements in the diagram,
30 is a terminal input/output program, 31 is a question interpretation program, 32 is a question execution program, 33 is a site management program, 34 is a communication control program, 35 is a registration table storage section, 36 is a secondary storage medium control device, 37 is a 2 It represents the next storage medium.

The registration table storage unit 35 in this embodiment stores the registration table 21 shown in FIG. There is.

First, registration of a database storage area to a site will be explained using this embodiment. 2 that stores the registration table 21, table data 13, and own database ID 22 in FIG.
When the next storage medium 37 is installed in the secondary storage medium control device 36 and a registration command for the database storage area is input from the terminal 4-1, the request is sent to the question interpretation program 31 via the terminal output program 30f. As a result of being recognized, a registration request is issued to the site management program 33. The site management program 33 is a secondary storage medium 37
Read the registration table 21 and own database ID 22 from
At the same time as setting in the registration table storage unit 35, the database ID
The database ID' is registered in the management table 20. Further, via the communication control program 34, the database ID and the site ID just registered and other sites are contacted.

On other sites, the database ID and site I
D' is set in the database ID management table of the own site. Registration is completed through the above processing, and the table data of site A becomes available.

Next, table data inquiries will be explained.

First, it is assumed that an inquiry command for processing tables A and B is input from the terminal 4-1. Also, the table name is the table creator name and the creator database I.
It is assumed that the data is expressed as a set of a DC database storage area name), an object name, and an object birth database ID.

The question interpretation program 31 is the terminal input/output program 30.
Interprets the command received via i and recognizes that the inquiry command targets table A and table Bi. Next, the registration table 21 stored in the registration table storage section 35 is scanned and compared with the storage database ID'i of table A and table B - own database ID22.

Based on the storage database IDt obtained as a result, the site management program 33 stores the site ID of the database ID.
t request. The site management program 33 scans the database ID management table and returns the site ID. As a result, it is determined that the storage site for table A is site A, that is, the own site, so the question execution program 32 performs processing within the own site. On the other hand, database I of table B
Since the storage site of D is found to be site B, that is, other than the own site, the site management table 10 is scanned to obtain the communication line address, and the communication control program 3 is sent to site B.
A processing request for table B is made via 4-money. In addition, if the table name does not exist in the registration table 21, the table name is based on the creator database III or the object a raw database ID (the database ID management table 20 is scanned, the site ID is obtained, and communication control is performed). Program 3
By making a search request for object storage sites to related sites via 4, it is possible to process table data existing in any site.

Even if the secondary storage medium 37 installed at site A is removed from the secondary storage medium control device 36 and installed at site C, if this embodiment is used, no contradiction will occur and the normal operation of the distributed database system can be guaranteed. Can be done.

〔Effect of the invention〕

According to the present invention, in a distributed database system,
Even if the storage medium that stores the database storage area that stores the table data group is installed at any site, or removed from any site and installed at any other site,
Since there is no inconsistency in the management of distributed table data groups, a new function called site transparency can be added to a single distributed database management system, and the availability for users can be completely improved. Furthermore, since it can be easily removed from the system when it is not needed, it can be expected to improve confidentiality.

[Brief explanation of drawings]

Figure 1 is an example of the configuration of a database management system using the site management method in the embodiment shown in Figure 4, Figure 2 (8) is an example of the configuration of a general distributed database management system, and Figure 3 is an example of the configuration of a general distributed database management system.
The figure is an explanatory diagram showing a conventional site management method in a distributed data management system, and FIG. 4 is an explanatory diagram showing an embodiment of the site management method according to the present invention. 10... Site management table, 11... Own site ID
. 13... Table data group, 20... Database ID management table, 21... Table registration table, 22...
... Own database ID, 33...Site management log■ 1 Figure Z Figure low 3 Figure 5-1 Te'7':-S 1z-7-71Sat low 4 0 Puito tZ Ren L Chi 2゛Le 1 Self-book I/) 3 Tefurteno-kun≠ =ρ Te°nita ='-ILD ¥1! Table: 1 Theme song humiliation 1k i2 Self-telephone base ep

Claims (1)

[Claims] 1. In a distributed database management system that handles a group of table data scattered on a system configured using multiple sites and a communication line network that connects them,
The site management method for identifying each site is based on a site ID (site identifier) unique to the site and a database ID unique to the storage area of the table data group existing at the site.
1. A site management method for a distributed database management system, comprising a site management device that controls a group of table data of each site by combining a (database identifier), a site ID, and a database ID.