JPH0863381A - Data base management method for distributed data processing system - Google Patents

Abstract

PURPOSE: To prevent the increase of loads of the system while holding the data consistency by surely writing data of the data base when write of data in a local data base is performed. CONSTITUTION: A data base management table 10 preserves an updating flag for each local data base as an index whether or not it is already written in a data base 7 when there is a writing and when writing is performed for each local data base as an index of whether or not the data base is changed for each table of a data base 4. When writing of data is performed by a data base 7 of local computer 2, writing of data of a data base 4 in a work station 1 is always performed. Thus, the contents of the data base 4 and that of the data base 7 become the same. When reading out the data base 7 of the computer 2, the local data base are read out after writing the local data base by the data base 4 only when the data base 7 and the data base 4 are different.

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 of a distributed data processing system suitable for managing a database of a distributed data processing system in which data is distributed and processed by a plurality of computers.

[0002]

2. Description of the Related Art Recently, along with the widespread use of personal computers, distributed data processing systems using personal computers or the like as local computers have become widespread.

FIG. 17 is a block diagram of a conventional distributed data processing system.

In the figure, 1 is a workstation as a host computer, 2 is a local computer such as a personal computer as a terminal computer, 3 is a network connecting the workstation 1 and the local computer 2, and the network 3 is
For example, a local area network (LAN) is used.

The workstation 1 is provided with a database 4, a database management processing section 5, and a transmission means 6. Further, each of the plurality of local computers 2 is provided with a local database 7, a local database management processing unit 8 and a transmission means 9.

With the above configuration, when the processing system is started up, the local database management processing unit 8 of each local computer 2-N requests the workstation 1 via the transmission means 9 to transmit the required database.

In the workstation 1, the database management processor 5 receives the request from the transmission means 6 and retrieves the requested database from the database 4 and sends it to the local computer 2-N via the transmission means 6.

In the local computer 2-N, the database is received via the transmission means 9, and the local database management processing unit 8 stores the required database in the local database 7.

Here, the relationship between the database 4 and each local database 7 will be described with a simple example. For example, as shown in FIG. 18, it is assumed that there is a database 4 of production costs of the A unit which is a part of the product in the database 4 of the workstation 1. The database 4 has 1) mechanical parts, 2) purchased parts, 3) processing cost, etc., and n) total manufacturing cost of A unit.

On the other hand, the local database 7 of each of the local computers 2-1 to 2-3 has the same data area as the above-mentioned database 4, and the local computer 2-1 mainly processes its own machine parts. The local computer 2-2 is mainly in charge of processing of purchased parts, the local computer 2-3 is mainly in charge of processing cost, and the local computer 2-N is mainly in charge of manufacturing cost of the whole product. In charge of calculation. In addition to the A unit database 4, B
There is a similar database 4 for units, C units, etc.

When the system starts up, the database 4 of FIG. 18 is downloaded to the local database 7 of the required local computer 2, and the data of the local database 7 corresponding to the data of the database 4 of FIG. 18, for example, the local computers 2-1 to 2-1. The data in the local database 7 of the local computer 2-3 match.

When the local database 7 of each local computer 2-N is downloaded from the workstation 1 in this way, each local computer 2-N is downloaded.
Each N accesses its own local database 7 to process various data.

As a result, the local computers 2-N share the data processing and execute the data processing to distribute the load and improve the efficiency of the entire system.

[0014]

However, as shown in FIG.
In the database management method of the distributed data processing system described with reference to FIG. 18, when the local database 7 is written in any one of the local computers 2-N, the data is exchanged with the local database 7 of another local computer 2-N. Differently, there is a problem that the consistency of the entire system cannot be maintained.

That is, when this is applied to the example described in FIG. 18, initially, the data in the database 4 in FIG. 18 and the corresponding data in each local database 7 match, but, for example, as shown in FIG. Local computer 2-
If 1 is calculated as the final result of calculating the manufacturing cost of the mechanical product, and it becomes \ a1, the corresponding area a of the local database 7 is written in a1, and the total is written in \ d1.

Further, if the local computer 2-2 discounts the purchased parts as shown in FIG. 19 and \ b becomes \ b2, the total \ d2 is written, and similarly, the local computer 2-
If the processing cost is changed and \ c is changed to \ c3, the total of 3 becomes d3, and the total \ d of the database 4 and the data between the local computer 2 are different. Therefore, the manufacturing cost of the A unit is not accurate, and if the B unit and the C unit are the same, the manufacturing cost of the entire unit is also inaccurate and unreliable.

To solve this, the local computer 2-
When N writes the data to the local database 7, the same data is written to the workstation 1, and when the local computer 2-N reads the data from its own local database 7, it downloads the database from the workstation 1 to download its own local data. It is conceivable to store the data in the database 7 and then read it.

However, own local database 7
If the data is downloaded from the workstation 1 every time the data is read from the workstation 1, the load on the local computer 2-N and the network 3 increases, and the processing efficiency of the entire system decreases.

Therefore, an object of the present invention is to provide a database management method for a distributed data processing system that maintains the consistency of the database of the distributed data processing system without increasing the load on the system.

[0020]

According to a first aspect of the present invention, a workstation for mounting a database and a plurality of local computers for mounting a local database are connected via a network, and the database of the workstation is stored in each local computer at startup. In the database management method of the distributed data processing system in which the data is downloaded to the local database, and each local computer reads the data in its own local database and processes the data, the workstation manages the correspondence between the database data and the local database for each local computer. The local computer has a database management table that sets the same status as the data to be stored.Each time the local computer writes the data to its own local database, it sends the data to the workstation. Every time the workstation receives the data and writes the data to the database, the different state of the database management table is set to be the same only for the local computer that has written the data to the database. Every time the data is read, a request for reading the corresponding data from the database is sent to the work station, and the workstation receives the database read request, determines the different status of the database management table, and responds only when the data is different. Data is transmitted, the different state is changed to the same, and the local computer receives the data and writes it in the local database.

According to a second aspect of the present invention, a workstation mounting a database and a plurality of local computers mounting the local database are connected by a network, and the database of the workstation is downloaded to the local database of each local computer at the time of startup, In a database management method of a distributed data processing system in which each local computer reads data from its own local database and processes the data, in a distributed data processing system, a workstation has a write permission flag for permitting writing to the data in the database and a local computer. Has a write management table for setting the different state of both the data in the database and the corresponding data in the local database, and each local computer writes work data every time it writes data to the local database. When the workstation receives a data write request and the data write request is received, the workstation notifies the write permission and writes the write management table when there is no write permission flag in the write management table. Change the setting of the permission flag to Yes to prohibit writing of data from other local computers, and change the setting of the write permission flag to Yes in the write management table when data writing is complete. Furthermore, the different state is set to be the same only for the local computer that has written the data in the database, and each time each local computer reads the data in its own local database, the corresponding data in the database is sent to the workstation. , The workstation receives the database read request, determines whether the write management table is in the same state, and sends the corresponding data only when they are different. As well as,
The difference state is changed to the same setting, and the local computer receives the data and writes it in the local database.

According to a third aspect of the present invention, a workstation that mounts a database and a plurality of local computers that mount the local database are connected by a network, and the database of the workstation is downloaded to the local database of each local computer at startup, In a database management method of a distributed data processing system in which each local computer reads data from its own local database and processes the data, a workstation permits writing corresponding to specific data in the database only to a specific local computer. The local computer has a terminal management table for setting the write permission setting and the different states of both the database data and the corresponding data in the local database for each local computer. Each time data is written to the workstation, a data write request is sent to the workstation, and when the workstation receives the data write request, the workstation notifies the write enable when the terminal management table is write enabled and When the writing of the data is completed, the same state of the terminal management table is set to be the same only for the local computer that has written the data to the database, and each local computer reads the work every time it reads the data of its own local database. The workstation sends a request to read the corresponding data from the database, the workstation receives the request to read the database, determines the different status of the terminal management table, sends the corresponding data only when they are different, and sends the different status. I changed the setting to the same and the local computer received the data and wrote it to the local database. That.

[0023]

According to the first aspect of the invention, when the data is written in the local database of the local computer, the writing of the data in the database of the workstation is always executed. This allows the contents of the database and all local databases to be the same. Moreover, when reading the local database of the local computer, the local database can be read after being written by the database of the workstation only if the local database and the database of the workstation are different. Therefore, it is not necessary to read the database of the workstation at all times when it is necessary to read the local database, and it is possible to avoid an increase in the load on the entire system while maintaining data consistency.

According to the second aspect of the present invention, when data is written in the local database of the local computer, the data in the database of the workstation is written, and the system load is maintained while maintaining the consistency of the data. In addition, avoiding the increase, and further, if there are a plurality of write requests subsequent to each other, the subsequent write is executed after the previous write is completed. This makes it possible to write data in a plurality of local databases to the database together and avoid inconsistency in data.

According to the third aspect of the present invention, it is possible to write to the database of the workstation within the local database of the local computer while avoiding an increase in the load of the system while maintaining the consistency of the data. It is possible to set the prohibition in advance, prohibit writing to the unwritable local database, and write to the writable local database only. Therefore, it is possible to prevent the data from being intentionally written in the database and destroyed.

[0026]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is an overall configuration diagram of a distributed data processing system showing a first embodiment of the present invention.

The same reference numerals as those in FIG. 17 showing the conventional example indicate the same or corresponding portions. In FIG. 1, a database management table 10 is additionally provided in the workstation 1 of FIG.
The configuration of the database management processing unit 5 is different from that of the database management processing unit 11, and the configuration of the local database management processing unit 8 of the local computer 2-N is different from that of the local database management processing unit 12.

Here, the database management table 10
Saves the update flag as an indicator of whether or not the database has been changed for each database table and as an indicator of whether or not it has already been read in the local database when writing is made for each local database. It is a table.

The database management processing unit 11 sets the presence / absence of writing in the database management table 10 to 1 when receiving data from the local computer 2 and writing the data in the database, and only the local computer 2 that has written the update flag. On the other hand, when the read request is made, the database data is transmitted and the update flag is set to 1 only when the update flag corresponding to the requesting local computer 2 is 0.

The local database management processing unit 12 is
Every time data is written to its own local database 7, a write request is sent to the database 4 and the data is transmitted.
When reading data from its own local database 7, a read request is issued for the data from the database 4, and the data is written to the local database 7 only when the data is received because it is different from the data in the database.

With the above configuration, the processing procedure shown in FIG. 2 and FIG. 3 will be described. First, when the processing system is started up, the local database management processing unit 12 of each local computer 2-N The transmission of the required database is requested to the workstation 1 via the transmission means 9.

In the workstation 1, the database management processing unit 11 takes out the requested database and sends it from the database 4 to the local computer 2-N via the transmission means 6 (201).

In the local computer 2-N, the database is received via the transmission means 9 and the local database management processing unit 12 stores the required database in the local database 7 (101).

In this way, when the workstation 1 downloads to the local database 7 of the local computer 2-N, each local computer 2-N accesses its own local database 7 to process various data. To do.

Thereafter, the local database management processing unit 12 performs the processing shown in FIG. First, when data is written in its own local database 7, the data is transmitted to the workstation 1 via the transmission means 9 (1
02, 103, 104).

On the other hand, the workstation 1 is
The data management processing unit 1 receives the data from the local computer 2-N via the transmission means 6 and performs the processing shown in FIG.
Done by 1.

Now, in the local computer 2-N, the database table name "000100" of the local computer 2-2.
When the data is written to the database 4 in "00", for example, "1" is set to the flag indicating the presence or absence of the writing of the database table name "00010000" in the database management table 10 shown in FIG. 4 (201, 202,
203).

Further, the update flag of its own local computer 2-2 is set to 1, while the other local computer 2-2 is set.
Reset the update flag of -N to 0 (204, 20)
5).

As a result, the database management table 10
Then, as shown in FIG. Two
The write flag "1" of "00010000" becomes the update flag "1" corresponding to the local computer 2-2, and the other update flags are reset to 0.

Next, when any of the local computers 2-N accesses the local database 7, for example,
When the local computer 2-3 accesses the database table name "00010000" of the local database 7, the local computer 2-3 sends a database read request to the workstation 1 (105).

On the other hand, the database management processor 11 of the workstation 1 refers to the database management table 10 and the write flag is 1 and the local computer 2-
When the update flag of 3 is 0, the update flag is set to 1 and then the database with the database table name “00010000” is downloaded and transmitted (206 to 21).
0).

As a result, the local computer 2-3 receives the download from the workstation 1 and reads it into its own local database 7 (106, 107, 10).
8).

On the other hand, when the write flag is 0, the corresponding database invariant is notified, and when the update flag of the corresponding local computer 2-N is 1, it is notified that the same database has already been read. (21,21
2).

Along with this, the local computer 2-3 can receive the database unchanged notification or the database read completion notification to access its own local database 7 (106, 107, 109, 110).

Further, for example, when the local computer 2-3 reads the local database 7 of itself again, the workstation 1 is requested to read it. In this case, in the database management table 10 shown in FIG. 5, the write flag of the database table name “00010000” is 1
The update flag of the local computer 2-3 is 1.

In this case, since the contents of the corresponding tables of the local database 7 of the local computer 2-3 and the database 4 of the workstation 1 are the same, even if a read request is made again, the read completion is already notified. . Therefore, there is no need to download again.

As described above, according to the first embodiment, when the data is written in the local database of the local computer, the writing of the data in the database of the workstation is always executed. As a result, the contents of the database and all local databases 7 can be made the same. Moreover, when reading the local database of the local computer, the local database can be read after being written by the database of the workstation only if the local database and the database of the workstation are different. Therefore, it is not necessary to read the database of the workstation at all times when it is necessary to read the local database, and it is possible to avoid an increase in load while maintaining data consistency.

FIG. 6 is an overall configuration diagram of a distributed data processing system showing a second embodiment of the present invention.

The same reference numerals as those in FIG. 17 showing the conventional example indicate the same or corresponding portions. In FIG. 6, a writing management table 15 is additionally provided in the local database 7 shown in FIG. Instead, a database management processing unit 16 is provided, and further, instead of the local database management processing unit 8, a local database management processing unit 17 is provided.
Is provided.

Here, the write management table 15 is set to 1 when the write permission flag for permitting (0) and disabling (1) the writing of data to the database 4 is written. It includes a built-in flag and a read flag for setting the local computer 2 of the same data as the data in the database to 1.

The database management processing unit 16 writes the database 4 when a data write request is issued from the local computer 2, sets the write permission flag and the write flag of the write management table 15 to 1, and writes the data. Only the read flag corresponding to the loaded local computer 2 is set to 1, and when the data is written to the database 4, the write enable flag is reset to 0, while when there is a read request, the read flag of the local computer 2 which is the read request. Only when 0 is 0, the data in the database is transmitted and the read flag is set to 1.

The local database management processor 17
When writing data to the local database 7, when a write request is made to the database 4 and a write enable notification is received, a read request is made when transmitting data and reading the data of the local database 7 The data is written to the local database 7 only when the data is received as different.

With the above configuration, referring to FIG. 7 and FIG. 8, first, when the system is started up, the local database management processing unit 17 of each local computer 2-N transmits the required database. A request is made to the workstation 1 via the transmission means 9.

In the workstation 1, the database management processing unit 16 takes out the requested database 4 and sends it to the local computer 2-N via the transmission means 6 (401).

In the local computer 2-N, the data is received via the transmission means 9 and the required data is stored in the local database 7 by the local database management processing unit 17 (301).

Thereafter, the local database management processing section 17 carries out the processing shown in FIG. First, when data is written in its own local database 7, a request for writing data in the local database is made (302, 30).
3).

On the other hand, in the database management processing unit 16 of the workstation 1, the local computer 2-N
A data write request from the server is received via the transmission means 6, and the processing shown in FIG.

Now, the write management table 15 is referred to for the write request to the database 4, and the write permission flag is set to 1.
It is determined whether it is 0 or not. For example, the database table name “10010000” of the writing management table 15 shown in FIG.
0 is set in the write permission flag of. In this case, the write permission notification is sent to the corresponding local computer 2-N (402, 403, 404).

Further, the write permission flag is set to 1 and
The write flag is set to 1 while the read flag is reset to 0 (405, 406, 407).

As a result, the write management table 15 is shown in FIG.
0, the database name, No. 12 "1001
The write enable flag 1 and the write flag 1 of "0000" are set, and the read flags are all reset to 0.

After that, when the writing of the local database 7 to the database 4 is completed, the write permission flag is set to 0 (304, 305) (408, 40).
9).

On the other hand, when the writing permission flag is 1 in the judgment 403, the writing prohibition is notified (403, 410). As a result, in the state of the write management table 15 shown in FIG. 10, the writing of the subsequent local database is prohibited until the writing of the preceding local database is completed.

Next, when any of the local computers 2-N reads its own local database 7, for example, when the local computer 2-1 reads the local database 7 having the database table name "10010000", the local computer 2 -1 makes a database read request (306).

On the other hand, when the write flag of the write management table 15 of the workstation 1 is 1 and the read flag of the local computer 2-1 is 0, the read flag is set to 1 and downloaded (411, 412, 413, 4
14, 415).

In the local computer 2-1, the download is input (306, 307, 308, 309). That is, the state of the flag shown in FIG. 11 is set.

On the other hand, when the write flag of the write management table 15 is 0, the database invariant is notified, and further,
Even if the write flag is 1 and the read flag is 1, it is notified that the same database has already been read (416, 41).
7).

Along with this, for example, the local computer 2
When -1 continues again, the database invariant notification and the database read completion notification are given and access becomes possible (30
7, 308, 310, 311).

According to the second embodiment, when data is written in the local database 7 of the local computer,
Data is written to the workstation database. At this time, if there are multiple write requests after the other party,
After the previous writing is completed, the later writing is executed.
This makes it possible to write data in a plurality of local databases to the database together and avoid data inconsistency.

FIG. 12 is an overall configuration diagram of a distributed data processing system showing a third embodiment of the present invention.

In the conventional example, the same reference numerals as those in FIG. 17 indicate the same or corresponding portions. In FIG. 12, a terminal management table 18 is additionally provided, and instead of the database management processing unit 5, the database management processing unit 19 is used as a local management unit. A local database management processing unit 20 is provided instead of the database management processing unit 8.

Here, the terminal management table 18 has a writing flag and a database which are set to 1 when the writing is made to the database 4 and the terminal ID which permits only the local computer 2 corresponding to the terminal ID to write the data. And a read flag for setting the local computer 2 of the same data to 1.

The database management processor 19 writes data to the database 4 when a write request is made from the terminal ID of the terminal management table 18 when a data write request is made from the local computer 2, and sets the write flag to 1 Set to
Only the read flag corresponding to the written local computer 2 is 1
When there is a read request, the data in the database is transmitted and the read flag is set to 1 only when the read flag of the local computer 2 of the read request is 0.

The local database management processor 20
When writing data to the local database 7, when a write request for the database 4 is made, and when a write enable notification is received,
The data is transmitted, written to the database 4, a read request is made when reading the data in the local database 7, and the data is written to the local database 7 only when the data is received because it is different from the data in the database.

The operation of the above configuration will be described with reference to FIGS. 13 and 14.
The local database management processor 20 of each local computer 2-N requests the workstation 1 via the transmission means 9 for the required database.

In the workstation 1, the database management processing section 19 takes out the requested database and sends it to the local computer 2-N via the transmission means 6 (601).

In the local computer 2-N, the database is received via the transmission means 9 and the local database management processing unit 20 stores the required database in the local database 7 (501).

Thereafter, the local database management processing section 20 performs the processing shown in FIG. First, when data is written to its own local database 7, a request for writing data to the local database 4 is made (50
2, 503).

On the other hand, in the database management processing unit 19 of the workstation 1, the local computer 2-N
The data write request from the database is received via the transmission means 6 and the processing shown in FIG. 14 is performed.

First, when there is a write request to the database 4, the database management processing unit 19 refers to the terminal management table 18 as shown in FIG. 15 and determines whether or not there is a terminal ID corresponding to the database table name. Is determined (602, 603).

For example, if the local computer 2-1 requests writing of data to the table name "20010000" of the database 4, the terminal management table 18
Then, the terminal ID 2-2 is preset. Therefore,
In this case, a write disable notification is given (604). This prohibits writing from other than the preset terminal ID.

On the other hand, for example, in the above case, the local computer 2-2 makes the table name “200100” of the database 4
When a write request of "00" is issued, the database table name "2001000" of the terminal management table 18 shown in FIG.
The terminal ID of "0" is 2-2. Therefore, a write enable notification is given, the write flag 1 is set, the read flag 1 is set, and writing to the database 4 is waited (605,
606, 607, 608).

Along with this, the local database management processing unit 20 of the local computer 2-2 uploads by writing notification (504, 505).

After that, one of the local computers 2-N reads it out to the local database 7. For example, if the local computer 2-1 uses the database table name “20010”
When the local database 7 of "000" is read, a database read request is made (609).

On the other hand, the database management processor 19 of the workstation 1 refers to the terminal management table 18. In this case, as shown in FIG. 16, the write flag 1
When the read flag is 0, the read flag is set to 1 and downloaded (610 to 613). In response to this, the local computer 2-1 inputs the download (507, 508, 509).

For example, the terminal management table 1 shown in FIG.
When the database read request for the database table name “20010000” is issued from the local computer 2-1 as in No. 8, the write flag is 1 and the read flag is 0. Therefore, the database 4 is changed, and the local database of the local computer 2-1 and the database 4 are different.

Therefore, in order to make the data in the local database of the local computer 2-1 and the data in the database 4 the same, the database 4 is read and downloaded.
As a result, the read flag of the local computer 2-1 is set to 1.

On the other hand, when the write flag is 0, the database invariant is notified, and when the write flag 1 and the read flag are 1, it is notified that the same database has already been read (614, 615). ).

According to the third embodiment, in the local database of the local computer, writing is possible or not possible is set in advance in the database of the workstation, and writing in the writing-disabled local database is prohibited. Only the writable local database can be written to the database. Therefore, it is possible to prevent the data from being intentionally written in the database and destroyed.

[0090]

As described above, according to the invention of claim 1, when data is written to the local database, the data of the database is always written, so that the contents of the database and all the local databases are written. Can be the same and the local database can be read after being written by the database only if the local database and the database are different. Therefore, it is not necessary to read the database at all times when it is necessary to read the local database, and it is possible to avoid an increase in system load while maintaining data consistency.

According to the second aspect of the invention, when the data is written in the local database, the data in the database is written. However, if a plurality of write requests are issued subsequently, the previous write is completed. Later writing is performed later. This makes it possible to write data in a plurality of local databases to the database together and avoid inconsistency in data.

According to the third aspect of the present invention, only the local database which is permitted to be written in the local database can be written in the database. Therefore, it is possible to prevent the data from being intentionally written in the database and destroyed.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram of a distributed data processing system showing a first embodiment of the present invention.

FIG. 2 is a flowchart showing a processing procedure of a local database management processing unit provided in the local computer of FIG.

FIG. 3 is a flowchart showing a processing procedure of a database management processing unit provided in the workstation of FIG.

FIG. 4 is an explanatory diagram showing an example of a database management table of FIG.

5 is an explanatory diagram showing another example of the database management table of FIG. 1. FIG.

FIG. 6 is an overall configuration diagram of a distributed data processing system showing a second embodiment of the present invention.

7 is a flowchart showing a processing procedure of a local database management processing unit provided in the local computer of FIG.

8 is a flowchart showing a processing procedure of a database management processing unit provided in the workstation of FIG.

9 is an explanatory diagram showing a first example of a write management table of FIG.

10 is an explanatory diagram showing a second example of the writing management table of FIG.

FIG. 11 is an explanatory diagram showing a third example of the writing management table of FIG.

FIG. 12 is an overall configuration diagram of a distributed data processing system showing a third embodiment of the present invention.

13 is a flowchart showing a processing procedure of a local database management processing unit provided in the local computer of FIG.

FIG. 14 is a flowchart showing a processing procedure of a database management processing unit included in the workstation of FIG.

15 is an explanatory diagram showing an example of a terminal management table of FIG.

16 is an explanatory diagram showing another example of the terminal management table of FIG.

FIG. 17 is an overall configuration diagram of a distributed data processing system showing a conventional example.

FIG. 18 is a first explanatory diagram showing a database management method.

FIG. 19 is a second explanatory diagram showing a database management method.

[Explanation of symbols]

 1 Workstation 2-N Local Computer 3 Network 4 Database 5 Database Management Processing Unit 6 Transmission Means 7 Local Database 8 Local Database Management Processing Means 9 Transmission Means 10 Database Management Table 11 Database Management Processing Unit 12 Local Database Management Processing Unit 15 Writing Management table 16 Database management processing unit 17 Local database management processing unit 18 Terminal management table 19 Database management processing unit 20 Local database management processing unit

Claims (3)

[Claims] 1. A workstation that implements a database and a plurality of local computers that implement a local database are connected by a network, and the database of the workstation is downloaded to the local database of each of the local computers at startup, and each of the local computers is downloaded. In a database management method of a distributed data processing system, wherein a computer reads data in its own local database and processes the data, the workstation stores both the data in the database and the corresponding data in the local database for each local computer. Each of the local computers sends the data to the workstation each time the local computer writes the data in its own local database. And, for each of the workstations writes to receive the data of the data into the database, de the difference state of the database management table to said database -
Data is written in the local computer, and each local computer sends a request for reading the corresponding data in the database to the work station every time it reads the data in its own local database. The station receives a database read request, determines the different status of the database management table, sends corresponding data only when different, and changes the setting of the different status to the same, and the local computer sets the data. A method for managing a database in a distributed data processing system, characterized by receiving and writing to a local database. 2. A workstation that mounts a database and a plurality of local computers that mount a local database are connected by a network, and the database of the workstation is downloaded to the local database of each of the local computers at startup, and each of the local computers is downloaded. In a database management method of a distributed data processing system, wherein a computer reads data from its own local database and performs data processing, wherein the workstation has a write permission flag for permitting writing to the data in the database and a local computer. Has a write management table for setting different states of both the data in the database and the corresponding data in the local database, and each of the local computers writes the data in the local database. Each time it inserts, it sends a request to write the data to the workstation, and when the workstation receives the request to write the data, the workstation writes if the write permission flag in the write management table is absent. Notified, the setting of the write enable flag in the write management table is changed to “no” to prohibit the writing of data from other local computers, and the writing is performed when the data writing is completed. The presence / absence of the write permission flag in the management table is changed to not present, and the different state is set to be the same only in the local computer that has written the data to the database. Every time the data in the database is read, a request for reading the corresponding data in the database is sent to the workstation, and the workstation sends the database read request. And Shin, said determining the difference status of the write management table, and transmits the data corresponding to only different time,
A database management method for a distributed data processing system, wherein the setting of the different state is changed to the same, and the local computer receives the data and writes the data in a local database. 3. A workstation that implements a database and a plurality of local computers that implement a local database are connected by a network, and the database of the workstation is downloaded to the local database of each of the local computers at startup, and each of the local computers is downloaded. In a database management method of a distributed data processing system, wherein a computer reads data from its own local database and processes the data, in the workstation, the workstation permits writing to specific data in the database only to a specific local computer. Including a terminal management table that sets the permission status and the different states of both the data of the database and the corresponding data of the local database for each local computer, and each of the local computers, Each time data is written to the local database, the data write request is transmitted to the workstation, and when the workstation receives the data write request, the workstation writes when the terminal management table indicates the write permission. When the writing of the data is completed and the writing of the data is completed, the different state of the database management table is set to be the same only in the local computer that has written the data in the database, and each local computer has its own local database. Each time the data is read, a request for reading the corresponding data in the database is sent to the work station, the workstation receives the database read request, determines the different status of the terminal management table, Send corresponding data only when they are different,
A database management method for a distributed data processing system, wherein the setting of the different state is changed to the same, and the local computer receives the data and writes the data in a local database.