KR930007231B1 - Database Access Method using Physical Address in DBMS of Electronic Switching System - Google Patents

Abstract

No content.

Description

Database Access Method using Physical Address in DBMS of Electronic Switching System

1 is a distributed structure diagram of a main memory DBMS having a distributed structure of an electron exchanger including the present invention.

2 is a diagram showing the configuration and message flow between S / W blocks for implementing a database access method using a physical address in a DBMS.

3 is a flow chart showing the overall flow of the database access method using the physical address of the present invention.

* Explanation of symbols for main parts of the drawings

ASS: Access Switching Subsystem

ASP: Access Switching Processor

CCS: Centran Control Subsystem DBBG: Data Base Backup Group

DBKG: Data Base Kernel Group

DBMS: Data Base Management System

DBSG: Data Base Supporting Group

INP: Interconnerction Network Processor

INS: Interconnerction Network Subsystem

LDB: Local Data Base, MMP: Man-Machine Processor

OMP: Operation and Maintenance Processor

The present invention relates to a database access method using a physical address in a DBMS (database management system) of an electronic switchboard.

The existing database management system's database approach was all done through database management information (directories, dictionaries). Therefore, the time taken to access the desired tuple actually depends on the database management information retrieval time and the start address of the relation to find information such as relation type and start address, attribute type and size, and access method. This is the sum of the time it takes to find the desired tuple. Therefore, this approach is not suitable for applications that are very sensitive to real time processing.

An object of the present invention is to provide a new approach that can minimize the time taken to access a database in a DBMS (Data Base Management System) of an electronic switchgear and enables real time processing.

The present invention is a method used by an application program with the strongest demand for real-time response in an electronic switchboard, and is a new approach in which an application program directly accesses a database using a physical address of the database. The present invention provides the physical address of a relation that an application wants to access in the initial stage of when the electronic switchgear starts to operate, and minimizes the time between search by allowing the application to directly search the relation using the physical address. The real-time processing function of the DBMS is firmly established, and no time burden is imposed on the real-time processing applications. In addition, although the access unit of the DBMS is a tuple, the access time of the attribute unit is also reduced to further reduce the data retrieval time.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings. 1 is a structural diagram of a main memory DBMS (Data Base Management System) having a distributed structure of an electronic switching system. The electronic switch is composed of three subsystems: an access switching subsystem (ASS), an interconnection network subsystem (INS), and a central control subsystem (CCS). ASS consists of Access Switching Processors (ASPs) that perform distributed call processing functions, INS consists of Interconnection Network Processors (INP) and Number Translation Processors (NTPs) that perform centralized call processing functions. It performs centralization functions related to operation and maintenance, and consists of Operation and Maintenance Processor (OMP) and Man-Machine Processor (MMP).

OMP's DBMS consists of three blocks: the Data Base Backup Group (DBBG), the Data Base Supporting Group (DBSG), and the Data Base Kernel Group (DBKG). The remaining processors consist of two blocks: DBSG and DBKG. DBSG performs communication function between DBMSs because user matching function and DBMS are distributed structures. DBBG performs the change control of duplicate data so that DBKG can change the duplicate data. DBKG actually retrieves, inserts, and deletes the database.

2 is a compile process in which EDMLP receives an application for accessing a DBMS from a VAX computer and generates information necessary for physical access of information required to access a database. This is a structure diagram showing the configuration and message flow between related S / W blocks for implementing an access method using a physical address in a processor of an exchange system.

The CHILL Embedded Data Manipulation Language Processor (EDMLP) creates a physical data structure (PDS) as part of the application during compilation for applications that want to use a DBMS. Generates the information needed to find the physical start address of the relation. Physical Data Structure The PDS is composed of a table type that contains the same number of tuples as the relation structure in the database and a pointer variable to the table. The information needed to find the physical start address of a relation is the address of a pointer variable that is needed to provide the application with the relation number and the physical start address of the relation obtained. The EDMLP sends the request analysis controller (RAC) of the OMP DBKG to find the physical starting address of the relation in the early stages of system operation. The transmitted information is the unique number of the relation to find the physical start address and the address of the PDS pointer variable.

The RAC of the OMP DBKG obtains the physical starting address of the relation from the database data in the LDB (Local Data Base), that is, the relation information directory, and stores it in the address of the PDS pointer variable provided by the EDMLP. As a result, the physical address of the relation necessary for accessing the database using the physical address is provided to the application program, and the application program accesses the DBMS to access the relation existing in the database by presenting the desired tuple number in the pointer variable generated by EDMLP. Without it, you can search as if you were using your own local variable. This process can be continued without access to the DBMS as long as the system is operating normally without losing the physical address of the relation.

Hi Table 1 shows an example of accessing a database using a physical address in a CHILL (CCITT High Level Language) application. In Table 1 below, (1) and (2) refer to PDSs generated by EDMLP, and the variable in which the physical address of the relation used by the application is stored is PhyAdr.

(3) is a command to limit access authority to processor PhyAces only by searching before accessing using physical address of relation R_xx. Such a function is essential to prevent a database from being destroyed or altered by a database accessor other than the DBMS, thereby destroying the integrity of the data. (4) is a data manipulation word that assigns the physical address of relation R_xx to the PhyAdr pointer variable created by EDMLP in the application. (5) is a CHILL statement that accesses directly without going through the DBMS by pointing tuple 0 as an index to the PhyAdr pointer containing the physical address of the relation R_xx obtained from the DBMS. Thus AttrVal will have the value of the first attribute of the first tuple of relation R_xx. (6) is a statement that increments the number of tuples by one to retrieve the second tuple, and (7) is a CHILL statement that retrieves the first attribute of the second tuple. Specifying the number of tuples here is not a problem because the application knows the contents of the data.

This physical approach is based on searching the entire tuple through the DBMS and using the desired tuple or accessing one tuple through the DBMS when the data access unit is an attribute (accessing one attribute to be used in a tuple at a time). The access time is much faster than that. Such characteristics are the core of the present invention.

TABLE 1

4 is a general flow diagram of accessing a database using a physical address of a DBMS. The DBMS of the electronic exchanger is a main memory DBMS having a distributed structure. EDMLP receives a database access request using a physical address from an application (1) and creates a PDS, a data structure for physical access (2). The PDS consists of a data structure of the same type as the relation structure existing in the database and a pointer to the structure of the relation. The EDMLP receives a request for access to a database from an application and generates information that restricts the search to only the requested application (3). The information needed to control database access rights is the process number of the application. This information is sent to the operating system at runtime and protects the database so that it can only be searched for applications that want to access the database using physical addresses. As a result, the integrity of the data does not occur when the database is destroyed or altered by an accessor other than the DBMS. In addition, EDMLP receives the request for the physical address of the relation to be accessed from the application and generates the information necessary for the DBMS to find the physical address (4). The information required by the DBMS to find the physical address of the relation is the relation number, the instruction number, and the address of the pointer for storing the found physical address (the address of the pointer variable that EDMLP already created in creating the physical data structure). The compiled result of the application program containing this information is loaded from the VAX computer into the electronic switchboard (5) and actually performed.

In the electronic switchboard, DBKG's RAC receives the request for the physical address of the relation that the application wants to access and starts the relation from the database management data RDIR (Relation Directory) using the number of the relation created by EDMLP. Find the address (6). RDIR contains the relation number, size, starting address of the tuple, the address of the table containing the attribute information, and the number of tuples. If the relation to be found does not exist in the database, report an error (7) and exit. If the relation to be found exists and the physical start address of the relation is found, the physical address of the relation obtained from the pointer variable included in the physical data structure is allocated (8). The application retrieves the attribute directly by pointing to the number of the tuple to look for in a pointer variable containing the physical start address of the relation obtained from the DBMS (9). The tuple number pointed to by a pointer variable represents the relative order of tuples within a relation. From this point on, the application can directly search for the desired data from the database without going through the DBMS. The process of obtaining the start address of the relation is done once at the start of the switchboard. After that, the application can directly search the database and search by attribute. Almost negligible. Therefore, such a data retrieval method can be used in an application that requires a very strong real-time processing, which is the most important development object of the present invention.

The present invention configured as described above operates in an electro-electric exchanger and has the following four effects. First, unlike the conventional DBMS, the database is accessed only once to get the physical address of the relation in the initial stage. During operation, the application can retrieve the data directly from the database. The time spent on is almost negligible. Second, it can effectively support applications that require the most real-time processing in data retrieval. Third, since the attribute-level access is possible, the time required for copying an attribute that is not needed can be reduced, which is more advantageous for real-time processing. Fourth, it is more effective when the data is large and the access method is sequential.

Claims (3)

A method of accessing a database using a virtual database in a data management system of an electronic switching system, comprising: a first step of creating a physical data structure in response to a request to access a database using a physical address in a data manipulation processor; A second step of generating information necessary for access right control by receiving an access right request for database access after the first step, and receiving a request for a physical start address of a relation to be accessed after the second step. A third step of generating information needed to find, after the third step a fourth step of finding the physical address of the relation and, if present, assigning the physical start address of the relation to a pointer variable of a physical data structure, after the fourth step Pointer side containing the physical start address of the given relation Point out the number of tuples, the application consists of a fifth step of retrieving the attributes directly from the database without going through the database management system to enable the real-time search of the database characterized in that New database approach used. 2. The database access method according to claim 1, wherein the physical data structure in the first step is composed of a table and a pointer variable to the same table as the relation structure in the database. The method of claim 1, wherein the method for searching for a specific attribute of a specific tuple in the fifth step is to indicate the number of tuples (relative order in a relation) as an address or variable in a pointer variable and use the attribute name. Featured database approach.