JPH03147031A - Data base access control system - Google Patents

Abstract

PURPOSE:To reduce access overhead by providing a means which holds mapping information in a mapping information control area. CONSTITUTION:When a data base operation request is issued from an application program 1, a data base control system 3 interprets the operation request by using data base operation information 2 and data base structure information 5. When the data base with respect to the request exists in a present node 8 as the result of analysis, an operation result is transferred to the application program 1 from a data base 10 in accordance with data structure information. When the data base with respect to the request exists in a different node 9, data base operation information 2 is transmitted to the other node 9 by using a communication line 12. The data control system 4 of the different node 9 executes the same operation as the data base control system 3 and transmits it to the opposite node 8. Then, the transmitted operation result is transferred to the application program 1.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to an access processing method for a database system, and in particular to reducing access overhead by managing mapping information generated by data conversion between schemas in a data manipulation language. This relates to a database access control method.

[Conventional technology]

Conventionally, in order to access a database, a data manipulation language provided by the database management system is written on an application program, compiled and executed using a language compiler.

Compiling a data manipulation language using this method
As described in Japanese Patent No. 2-164136, data manipulation information in which logical structure information of a database is embedded is expanded into an application program object. A database management system receives data manipulation information from an application program, performs conversion processing from logical structure to physical structure, generates mapping information so that the physical structure information can be referenced most efficiently in order to handle the database, and then converts the data accordingly. Perform database access.

In the case of distributed database access, conversion processing is similarly performed between the database structure that summarizes the database structure of each note and the database structure that exists in each individual node. In this process, mapping information is generated, and database access is executed accordingly.

[Problem to be solved by the invention]

In the above conventional technology, the data operation information issued from the application program holds only logical database information, so in order to access the database,
In the access process, it is necessary to perform inter-schema conversion up to the physical database structure. Also, when accessing a distributed database, conversion between schemas is performed from the global logical structure to the target physical structure of the access target node. Mapping information created in the process of converting between these schemas is deleted after access is completed. Therefore, even if the same data manipulation language was reissued, conversion processing between schemas was necessary.

Furthermore, in large-scale online systems, it is common for the same application program to handle different data.
Although the data manipulation language issued by each application program is the same, the database management system performs schema-to-schema conversion processing as an extension of each application program, resulting in redundant processing. Furthermore, among data manipulation languages, only general search-related operation statements are capable of complex descriptions, and for other operation statements, the description pattern of the operation statement can be easily predicted, and mapping information can be generated in advance. , no pre-generation is performed.

The present invention has been made for the purpose of providing a database access control method that reduces the above-mentioned access overhead as much as possible.

[Means to solve the problem]

The system of the present invention consists of a plurality of nodes having processors and databases, and a communication network that interconnects each node, and each node has an application program that issues database access requests and receives requests from the application program. A database management system that accesses the data bus or another notebook database and a data dictionary/directory that manages meta information of the database structure and database structure information that defines the database structure and that is written in the application program. In a database processing system consisting of a compiler that compiles a data manipulation language, there is a means for generating mapping information that converts the data manipulation language from a logical structure to a physical structure, and the generated mapping information is stored in a mapping information management area. A means for reusing the retained mapping information when reissuing the same data manipulation request from an application program, and a means for reusing the mapping information that is maintained when the same data manipulation request is reissued from the application program, and a means for reusing the data manipulation information managed by the data dictionary in the database manipulation information when compiling the application program. This is achieved by reusing mapping information and making it shareable among multiple application programs by means of embedding sentence identification information and means of managing a mapping information management area using the identification information. In addition, in the case of a data manipulation language sent from another node, the means for adding the management information in the mapping information management area to the result information and returning it to the sender, and the means for registering the same management information, are the same again. When transmitting data manipulation language, provide a means to resend registered management information.

Furthermore, for data manipulation languages that are used as standard, a means is provided to generate mapping information in the database structure information in advance.

[Effect]

In the database access control method according to the present invention, by providing means for holding mapping information in a mapping information management area, conversion processing associated with logical-physical structure conversion of the same data manipulation language is eliminated. It can also be shared by multiple application programs, reducing memory space.

Furthermore, by registering the management information in the mapping information management area of the other party, if the same data manipulation language is to be retransmitted to a registered node, the management information is sent without sending the data manipulation language. In this case, since the amount of management information is small, the amount of messages to be sent can be reduced. Furthermore, the receiving party can directly search for mapping information from the transmitted management information.

Furthermore, since there is no need to convert the standard description pattern data manipulation language into mapping information or to manage it in the mapping information management area, higher-speed processing becomes possible.

〔Example〕

Embodiments of the present invention will be described below based on the drawings.

FIG. 1 is an overall configuration diagram of a database system implementing the present invention. In the figure, 1 is a user application program, 2 is database operation information generated as a result of compiling the database operation language written in the application program 1 with a language compiler, 3 and 4 are database management systems, 8 and 9 are nodes, 1 o.

Reference numeral 11 indicates a database, and reference numeral 12 indicates a communication line connecting the nodes 8 and 9.

7 is a mapping information management area that can be shared by a plurality of application programs, and 6 is mapping information for standard database operation information prepared for each operation target defined in the database structure information 5. Here, the database structure information includes information regarding a logical database structure and a physical database structure, and is managed by a database management system in a format in which the logical and physical are associated.

The outline of the database access processing procedure in the database system of this embodiment is as follows.

When a database operation request is issued from the application program 1, the database management system 3 takes the database operation information 2 and the parameter group passed along with the issuance request, and requests the operation using the database operation information 2 and database structure information 5. Interpret and execute. As a result of the interpretation, if the database 1 corresponding to the request exists in the local node 8, the data is retrieved from the database 10,
The operation result is transferred to the application program 1 according to the database structure information 5.

If the result of the interpretation is that the database corresponding to the request exists in another node 9, the database operation information 2 is sent to the other node 9 using the communication line 12. The data management system 4 of the other node 9 uses the received data to execute the same operations as the database management system 3, and transmits the execution results to the other node 8. The transmitted operation result is transferred to the application program 1.

Next, the present invention will be explained using FIGS. 2 to 5.

FIG. 2 is a schematic explanatory diagram of the process up to the generation of database operation information. A database operation language 21 written in an application program 20 is translated by a compiler 22. The definition specifications of the database described below are JIS: X3004-1987. Based on the database language NDL.

The database operation language 21 is analyzed by the compiler 22 through processes such as lexical analysis and syntactic analysis, and then the database operation language 2 is stored in the data dictionary 25.
It is determined whether the item handled by is actually defined in the database.

The database definition information in the data dictionary 25 is information obtained by W9ing the database definition 26 defined by the user using the definition utility 27, and is information regarding the database structure as shown in FIGS. 3, 4, and 5. .

Figure 3 shows the subschema, and Figure 4 shows the logical schema.

FIG. 5 shows the storage schema.

A logical schema is defined by records, sets. The record is made up of a plurality of components, each of which is itself the smallest piece of logical information. A set consists of multiple records and defines parent-child relationships.

A subschema is a part extracted from a logical schema that is used for each application program, and a storage schema defines the storage location for the logical schema and the access method using methods such as hashing and indexing. It is something that

Hereinafter, when we refer to a logical structure, we refer to subschemas and schemas, and when we refer to it as a physical structure, we refer to a storage schema.

In FIG. 3, records LA, r, and B of the logical schema are extracted as record views VA and VB, and their constituent elements are LA-L as constituent element views.

LA2. LA3. LBI, LB2 is VAI, VA2゜V
A3. VBI and VB2 have been redefined.

Data manipulation language 21 is VA i-= 1 ga VA2=
2 and VA3='A'.

As mentioned above, the compiler 22 converts the data manipulation language 21 into l? The database operation information 24 is developed in the application program object 23. The database operation language 21 is embedded in the database operation information as operation information in a binary format.

Here, DML (Data Manipulation Language)
The identification P is obtained by registering the data manipulation language in the data dictionary at the stage of compiling the application program, and adding a unique number to each data manipulation language unit. If data manipulation languages of the same format are registered at the time of registration, their DML identifiers will be the same.

In the case of a database operation language other than the database operation language 21, which is not used to perform somewhat complex database operations, only the DML identifier is output to the database operation information 24. In that case, mapping information, which will be described later, is created in the corresponding database structure information 29 and output to the directory 28. The database structure information 29 is a table called a directory in which database definition information of logical schemas and subschemas is converted into binary code and converted into a format that is easy for computers to refer to.

FIG. 6 is an overview of database access processing of the database management system in the own node 8.

When the database management system 3.4 receives a database operation request from the application program l or another node (step 31), in step 32 it determines which node will process the database operation based on the database operation information 2, parameters, and database structure information 5. Determine whether If the corresponding database operation request is processed by the own node, it is determined in step 33 whether mapping information has been generated.

The mapping information converts the logical structure information of the database specified in the database operation information to the physical structure information in the database structure information 5.

This is intermediate information for efficient processing according to the physical structure when actually accessing the database.

If no mapping information has been generated in step 33, mapping information is generated in step 34. The details will be explained later using FIG. 8.

If mapping information has already been generated in step 33, mapping information on the mapping information management area is searched for using the DML identifier in step 35.

Once the mapping information is ready, step 36
Performs database access processing.

If it is determined in step 32 that the processing will be performed by another node, it is determined in step 37 whether there is partner node mapping management information.

Partner node mapping management information is created when a database operation request is determined to be processed on another node and executed on another node, and is managed in the mapping information management area of that node. After the mapping information is processed in another node, it is sent back as result information, and the returned mapping management information is registered in the mapping information management area of the own node. If there is no partner mapping management information in step 37, data manipulation information 2 and parameter information are edited and transferred to the partner node as a data manipulation request to the partner node in step 38. If there is partner node mapping management information, the mapping management information and parameter information are edited and transferred to the partner node.

In step 4o, the result information is received from the other node and transferred to the own node application program 1. Among the received results, if data manipulation information 2 was sent at the time of transmission, the other node mapping information management information will have been returned, so the corresponding mapping information in step 41 has already been registered in the mapping information management area. If so, the corresponding node mapping management information is registered in the mapping information management area in step 42.

Next, FIG. 7 will be explained. FIG. 7 shows step 38.
39 is an outline of the process performed by the other node when a database operation request is transferred to the other node in step 39.

The partner node database process in step 60 is started when the message sent in steps 38 and 39 is received in the message reception process in step 61. When the message is received, the database operation information 2 in the message is extracted. Since what exists in the database operation information sent to another node is an overall logical database structure that is common among nodes, it is necessary to convert it into the logical database structure possessed by the receiving node.

In step 62, conversion processing is performed from this overall logical structure to a logical database structure possessed by each node.

At this time, similar to mapping information by logical-physical conversion, mapping information from the overall structure to individual nodes is generated. This mapping information is also managed within the mapping information management area.

The self-node database access process of step 63 is the same as the database access process (step 30) in FIG. At this point, processing is performed using the database operation information converted in step 62.

When step 63 is completed, the access result is edited as a result message in step 64.

In step 65, it is determined whether the mapping information of the database operation information for which the database access process has just been performed was generated during the process of the current database access process. If generated in the current process, step 6
In step 6, the management information (location information) of the mapping information management area in which the mapping information is registered is added to and edited in the message created in step 64, and in step 67, the edited return message is sent to the sender. This partner node database processing shown in FIG. 7 makes it possible to access a database located in another node.

Also, if the same database operation information is sent again,
Since the mapping information has already been created, this processing step can be omitted and the processing speed can be increased.

Next, FIG. 8 will be explained. FIG. 8 corresponds to step 34 in FIG. This process generates mapping information, which is conversion information that allows high-speed conversion from a logical structure to a physical structure or vice versa.

In step 81, the application program 1 determines whether the database operation information 2 for which the database operation request was issued is a database access command.

At step 82.90, the type of database operation request is determined. At step 82, the type of operation request is search-related, and at step 90, the type of operation request for addition/update is determined, and control is transferred to each process.

In the search process, a series of processes from steps 84 to 89 are performed. Search processing is usually used in batch processing, which searches the entire database to obtain statistical information, and online processing, which searches for specific data in the database by specifying specific conditions. Can be divided.

In step 84, it is determined whether a conditional expression for selecting data in the database is specified in the search operation. If a conditional expression is specified, key information within the conditional expression is extracted.

Key information is an access method such as an index or hashing added to access the record at high speed.
In order to use them, certain components within the record are
This is an access means to achieve high-speed access created using the unique property within a record. If these are specified in the conditional expression, high-speed access using that key information is possible, so the key information is extracted from the conditional expression. If there is a key condition and the key is an index key, a command word for performing a binary search on the index is generated. In the case of a hash key, when using system standard hashing processing, a hashing processing instruction Jc using that key is generated.

If there is no key condition or if there is a clause other than the key condition expression, the contents of the constituent elements on the record are generated into a group of command words to be determined according to the condition expression.

In step 87, a database access method is selected from the presence or absence of a key condition, an indicator indicating a specific data position in the current pig base, and the like. In step 88, a command word for transferring the retrieved data to the application program 1 is generated. This is because the arrangement of the components provided to the application program and the components in the database 2 do not match, or the data attributes of the components are different, so component replacement and attribute conversion are performed.

In step 89, information such as information for associating logical records with physical records, information regarding other related records through database operations, etc. is created. These mapping information are created on the mapping information management area. The mapping information management area consists of identification information uniquely added to the database operation information for each application, identification information of the issuing module, and a pointer to the mapping information.The mapping information can be shared between applications that use the same data operation information. It is possible to share.

Step 90 is a process for determining whether it is an addition or update type operation, and if it is determined that it is an addition or update type operation statement,
It is determined in step 91 whether or not it is a standard operation. Addition and update operations generally only add or update one record, and add a specific component in the operation statement.

It is not common to perform update processing. Therefore, the update and addition operation statements written in the application program 1 are written using the simplest description possible with the operation statements. This is defined as the standard form of update and addition operation statements. This mark 19H type is specified for each record specified in the view used by the application program 1, so its mapping information can be created in the database structure information created for each view. be.

Step 92 shows a process of speeding up the mapping information generation process using the mapping information created in advance.

〔Effect of the invention〕

According to the present invention, it is no longer necessary to generate mapping information, which has conventionally been generated for each database operation request, after the mapping information is generated at the time of the first operation request. Furthermore, since it can be shared between multiple applications, in cases where data handled by the same application program is different, such as in a large-scale online system, there is no need to generate mapping information as an extension of each application program.

Furthermore, since mapping information is generated in advance in the database structure information for standard operation statements, mapping information generation is completely unnecessary.

When dealing with distributed databases, in addition to the above effects, the amount of messages to be forwarded can be reduced.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing an overview of the database system according to the embodiment of the present invention, Fig. 2 is an explanatory diagram of the process of generating database operation information, Fig. 3 is an illustration of a subschema, and Fig. 4 is an illustration of a logical schema. 5 is an illustration of a storage schema, FIG. 6 is a processing flow diagram showing the process of database access in an embodiment of the present invention, and FIG. 7 is an embodiment of accessing a distributed database in the present invention. FIG. 8 is a processing flow diagram showing the process of mapping information generation according to the embodiment of the present invention. 1... Application program, 2... Database operation information, 3, 4... Database management system, 5... Database structure information, 6... Mapping information for standard database operation information, 7... Mapping Information management area, 8, 9...nodes, 10, 11 database. Tora Ward Tomi 4 Zufuzu No.)

Claims (1)

[Claims] 1. Consisting of a plurality of nodes having processors and databases and a communication network interconnecting each node, each node has an application program that issues a database access request and an application program that compiles the application program. A data dictionary/compiler that manages meta information such as a compiler and database structure
In a database processing system consisting of a directory and a database management system that accesses the own node database or another node database in response to requests from application programs, the data manipulation language is changed from a logical structure to a physical structure in order to access the database. A database access control method characterized by retaining mapping information to be converted in a mapping information management area and reusing it. 2. The database access control system according to claim 1, wherein database operation statement identification information is embedded in the database operation information, and mapping information is managed in a mapping information management area using this information. 3. A database access control method characterized by creating mapping information in advance in the database structure information for a data manipulation language that is used as a standard. 4. In the case of database operation information sent from another node, the management information in the mapping information management area of the own node is sent to the source node along with the database access results, and if the same database operation is performed again, the management information is sent to the other node. A database access control method characterized by sending.