JPS6055854B2 - Data base storage area management control method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides a data base storage area management control method;
Particularly in a data processing system that has a database, the data base management system should store 1
The present invention relates to a database storage area management control system that automatically manages the storage location of two records and allows the management process to be performed efficiently and at high speed.

Records stored in a database need to be stored in such a way that they can be accessed efficiently, keeping in mind not only individual records, but also mutual relationships, that is, set relationships, with other records.

That is, for example, when storing records related to personnel management, the records of a certain individual should be stored in the same block as the records of other people in the same department as much as possible.
It is desirable to be able to access only a single device for processing based on departmental units. In such record storage locations, it is extremely troublesome for the user to be aware of and use the application program to consider each physical medium, ie, device, in which records should be stored. For this reason, database management technology that allows application programs to perform logical processing without being aware of the physical medium has been adopted. , determining the storage location in the event of an overflow has been left to the logic of each individual application program.

In other words, when trying to store records with the above points in mind, it is possible to check in the user management table, etc. which block storage positions are free, and which record storage positions within one block storage position are free. It was necessary to manage whether or not there was a record, and to specify the storage location by referring to the contents of the user management table when storing a certain record. SUMMARY OF THE INVENTION The present invention aims to solve the above-mentioned problems, and aims to make the area management of a database independent of application programs and to be performed by a database management system.

Therefore, the database storage area management control method of the present invention is provided with a database that stores resources necessary for processing an application program in units of records, and stores the records in response to a record storage request from the application program. In a data processing system having a processing function for storing records on the database, the type of record to be stored is specified by the application program. A database management system is provided that executes a process of storing records in a position on the database corresponding to the type of record, and the storage position on the database is composed of a plurality of record storage positions. To deal with the shortage of free record storage positions within one block storage position during the storage process of the next record following the previously stored record when the records are divided into multiple block positions and stored one after the other. An overflow area is defined on the above-mentioned data base, and the above-mentioned data base management system includes a free first block number holding unit that manages free block storage positions among the plurality of block storage positions, and an overflow area for each of the above-mentioned blocks. an empty byte position holding unit that is created corresponding to the storage position and manages an empty state record storage position within the block storage position;
A latest processed record address information holding unit that manages the record storage position by referring to the previously stored record storage position when storing records belonging to the same type one after the other, and a free space belonging to the overflow area. Based on the contents of the empty first block number holding section in the overflow area that manages the storage position of the status block,
The database management system is configured to determine and store the record that has been requested to be stored from the application program by determining the storage position of the next storage record within the one block storage position, and further, the database management system first stores the record that has been requested to be stored from the application program. - Determine the storage position of the next record based on the contents of the address information holding section, and if the above determination cannot be made based on the contents of the latest processed record/address information holding section,
The next record storage position is determined and stored based on the contents of the free first block number holding section, the contents of the free byte position holding section, and the contents of the overflow area free first block number holding section. There is.
This will be explained below with reference to the drawings. FIG. 1 shows the configuration of an embodiment of the present invention at the level of program processing, and FIG. 2 shows an explanatory diagram illustrating the processing operation for determining the storage position of the next record to be stored next by the database management system. . In Figure 1, 1 is an application program, 2 is data●
Base management system, 3 is data ● base processing system, 4 is area management function, 5 is data base, 6 is data ● dictionary ● and ● directory (hereinafter) 1) D
/D) in which the structure of the database is defined such as the type of record, the relationship between the types, the size of the area, and information regarding the physical medium, etc., are collectively managed. In the case of the present invention, the database management system 2 manages and executes record storage processing for the database 5 by itself.

That is, when the application program 1 stores a certain record, the application program only needs to specify the type of record and issue a write command. For example, regarding personnel management work, when storing records of a certain individual, if the individual belongs to a certain department, for example, department A,
It is sufficient to issue a write command by indicating that the record to be stored is a personal record belonging to department A. When the application program issues a record storage request to the database management system as described above, the database management system 2 uses the area management function 4 to determine the optimum storage location by referring to the DD/D 6.

Based on the determined next storage record storage position, the database management system 2 stores the data base 5.
Stores to . In determining the optimal location, the database management system 2 stores (1) records of individuals belonging to the same department in the same block on the same physical medium as much as possible; Based on which block storage position is empty among the plurality of block storage positions configured on the base 2, and which record storage position among the selected free block storage positions is empty, etc. It may be considered that the above-mentioned position determination is performed by FIG. 2 shows an explanatory diagram for explaining the processing operation for determining the storage position of the next storage record.

The reference numeral 5'' in the figure conceptually represents the storage area space on the database shown in Figure 1, 7 the storage area space,
8 is an overflow area space, 9 is a block storage position, 10 is a record storage position, and 11 is a header section provided at the beginning of each block storage position, which is a system
What is called a record, 12 is a latest processing record/address information holding section, 13 is an empty first block number holding section, 14 is an overflow area free first block number holding section, 15 is an overflow extraction block number holding section, 1
6 represents an empty byte position holding section. The latest processed record/address information holding unit 12 is provided in the upper &DD/D6, and stores the address information of the record storage position of the record stored earlier in the process of storing records of the same type one after the other. is created to speed up the positioning process when storing subsequent records.

Note that the holding section 12 is provided for each type of record, and records that have a parent-child relationship, such as a record of department A and a record of an individual belonging to department A, are handled as records belonging to one type. The empty head block number holding unit 13 is provided, for example, in the DD/D 6, and is located at empty block storage positions 9-1, 9-2, etc. belonging to the storage area space 7. Holds the first block number of the state.

When all of the storage area space 7 is empty, the contents of the holding unit 13 indicate the block storage position 9-1, and records belonging to one type are stored in the block storage position 9-1. In this case, the block storage position 9-2 is designated to store records belonging to other types in the next block storage position 9-2. The free first block number holding unit 14 in the overflow area stores, for example, the above-mentioned DD.
/Can be earned within D6. As mentioned above, when processing records belonging to the same type to be stored one after the other, that is, when attempting to store multiple records belonging to the same type together in the same block storage position as much as possible, the block storage There may be a shortage in position.
For this purpose, an overflow area space 8 is defined as a so-called subnote of the block storage position. The empty first block number holding unit 14 in the overflow area stores block storage positions 9-5, . . . belonging to the space 8.
. . . holds the first block number of one that is free. When all of the overflow area space 8 is empty, the contents of the holding section 14 point to block storage locations 9-5. The contents will be updated sequentially below. The overflow cutout block number holding unit 15 stores each block storage position 9-1, 9-2, . . .
... are provided in the header section 11 of each.

Then, when an overflow occurs when records belonging to the same type are sequentially stored in the block storage position as described above, it is specified which block storage position in the overflow area space 8 is to be used as the subnote. . In general, it can be considered that the contents of the free top block number holding section 14 in the overflow area are set. The intra-block relative empty byte position holding section 16 is provided in the header section 11 of each block storage position 9-1, 9-2, . . . .

And record storage position 10 within the block storage position
Indicates how full the area is. That is, it indicates the position of the first empty byte within the block storage position. Now, as in the conventional case, the user has to decide, regarding the storage area space on the database, (1) which block storage position to which block storage position should be secured as the storage area space 7 shown in FIG. Ii) Register in advance which block storage position to which block storage position is to be secured as the overflow area space 8 shown in FIG.

That is, the registration is held within the DD/D6. There are the following departments, A, B, C, etc., and the explanation will be given assuming that personal records belonging to each department are stored.

Let us consider the case where the record of individual A1 belonging to the first division A is stored. In this case, the database management system 2 shown in FIG. 1 uses the area management function 4 to determine the storage location as follows. That is, (1) the content A (A) of the most recently processed record/address information holding section 12 existing in the DD/D6 is checked.

Note that content A (A) indicates an empty byte position within a certain block storage position 9, as will be described later, when records 5 belonging to the same type are stored one after the other. At the above point in time, it is not possible to determine the storage location of the target block based on the content A! (
This is the first individual record belonging to department A). (2
) Therefore, the target block storage position is determined based on the contents B of the free first block number holding section 13 in the DD/D6.

In this case, the content 3B indicates the block storage position 9-1. (3) After determining the block storage position 9-1, check the header part of the block storage position 9-1 and determine the next record storage position based on the content E of the intra-block relative free byte position holding section 16. and store record A1. (4) The above content A (A) is updated to indicate the block storage position 9-1 and the first byte of the storage position of the record A2 shown in FIG. 2 within the block storage position 9-1, and the above content A is also updated to point to the first byte of the storage position of record A2.

(5) The size of one record, for example, record A1, is undefined, and even when record A1 is stored, it may not be possible to store it in the block storage position 9-1.

In this case, the content B of the holding unit 13 is updated to (B+1) (in the illustrated case, block storage position 9-2 is specified), and the next block storage position 9-2 is determined as the target block storage position. . Then, the processes following the above process (3) are performed. Next, consider storing the record of individual A2.

In this case, it is determined as follows. That is, (6) DD/D
The content A (A) of the most recently processed record/address information holding unit 12 existing in 6 is examined. At this point, since the record of another individual A1 belonging to department A has already been stored, the block is blocked based on the above content A (A) corresponding to the type of record (in this case, department A). Storage position 9-
1 is determined. (7) Next, according to the contents E of the holding section 16 in the header section of the block storage position 9-1,
The next record storage position is determined and record A2 is stored.

(8) Then, the above content A (A) and content (E) are updated.

Records A3 and A4 are stored by the same processes as the above processes (6), (7), and (8).

However, individuals B1 to B8 belonging to Illustration Department Section B in Figure 2
As in the case of trying to store the record A, the block storage position 9-2 in which the record B1 was initially stored may be insufficient. Record Otsu 1, Otsu 2...
..., Otsu 5 is stored in the same manner as in the above processes (1) to (8), but it goes without saying that in this case, content A (Otsu) is used as the content in the holding unit 12.

When trying to store record O 6, the block storage position 9-2 becomes full. In this case, the process is as follows. That is, 9) the header section of the block storage position 9-2 is examined and the target block storage position is determined based on the content D of the holding section 15;

In the case shown in FIG. 2, since this is the first overflow, the above content D is an initial value, and in this case, the above-mentioned holding unit 14
Based on the content C of , block storage position 9-5 is determined as the target block storage position. 01 and set the above content C to the above content D, and set the above content C to (C+1
). Processes similar to the above processes (7) and (8) are then performed. As explained above, in the case of the present invention, the database management system 2 uses the area management function 4 to automatically determine the next record storage position. When records belonging to the same type are stored, the block storage position where the previously stored record was stored is automatically found, and the later-described record is stored in the same position. Therefore, it becomes possible to move to an optimal position in consideration of access efficiency, and the burden on the application program side is greatly reduced. Furthermore, in determining the storage location, the most recently processed record/address information holding unit 12
If the position cannot be determined, the contents of the empty head block number holding section 13 are checked first. Therefore, the processing speed is improved when records belonging to the same type are stored one after another. Particularly in relation to update processing, it may be more convenient for the empty head block number holding section 13 to exist in the header section of the storage area space 7, like the overflow cutout block number holding section 15 and the like. In such a case, when checking the contents of the holding unit 13, the data
Whereas it was necessary to access the base 5, by providing the most recently processed record address 12, the processing speed is greatly improved.

[Brief explanation of drawings]

FIG. 1 shows the configuration of an embodiment of the present invention at the level of program processing, and FIG. 2 shows an explanatory diagram illustrating the processing operation for determining the storage position of the next record to be stored next by the database management system. .

Claims (1)

[Claims] 1. Data processing that is equipped with a database that stores resources necessary for the processing of an application program on a record-by-record basis, and has a processing function that stores records on the database in response to record storage requests from the application program. The system includes a database management system that executes a process of specifying the type of record to be stored from the application program and storing the record in a position on the database corresponding to the type of record. , when the storage location on the database is divided into multiple block locations each consisting of multiple record storage locations, and when the next record is stored following the previously stored record, An overflow area is defined on the database in order to cope with the shortage of free record storage locations within the one block storage location in the data base.
The base management system includes a free first block number holding unit that manages free block storage positions among the plurality of block storage positions, and a free start block number storage unit that is provided corresponding to each block storage position and stores free state records within the block storage position. Free byte position holding unit that manages the position, and latest processed record address information that manages the record storage position based on the previously stored record storage position when processing records belonging to the same type to be stored back and forth. Based on the respective contents of the holding unit and the free first block number holding unit in the overflow area that manages the storage position of free blocks belonging to the overflow area, records requested to be stored from the application program are stored in the one block mentioned above. The database management system is configured to determine and store the storage position of the next storage record within the storage location, and further, the database management system first stores the next record based on the contents of the latest processed record address information holding unit. If the storage position cannot be determined based on the contents of the latest processing record/address information holding section, the contents of the free first block number holding section, the contents of the free byte position holding section, and the above overflow area are determined. A database storage area management control method characterized in that the storage position of the next record is determined and stored based on the contents of a free head block number holding section.