KR100302899B1 - The method of loading relation data in on-line while using database - Google Patents

Abstract

The present invention relates to a method for loading relation data online from a database in operation, and to create a database file (relation file) to load the damaged part in the event of a system failure due to a database corruption by using a database production tool. By loading only those relations into the system memory, you can prevent service interruptions and maintain a stable system, recover only the damaged parts quickly, and in the case of redundant relations, they are automatically loaded into all subsystems where they are loaded. In addition to reducing the database inconsistency of the system, the use of the loading function ensures the reliability of the system and has the advantage of ensuring convenient and stable system operation.

Description

The method of loading relation data in on-line while using database}

The present invention relates to a method of loading relation data online from a database in operation, by regenerating the damaged parts individually and loading them directly into a memory corresponding relation area of the exchange system, without disrupting service of the system due to database corruption. It is about how to recover data.

In general, electronic switchboards have the functions of diagnosing and finding inconsistencies in the system when an unexpected abnormality occurs during operation, and recovering when an inconsistency or damage is found.

Among them, when the database stored in the disk is damaged, that is, when the relation value of the exchange database is damaged, there is a method of recovering using a system query or using an operator command.

However, in the above method, if there are many tuples of relations to be restored or if these relations are redundant relations, they cannot be changed or restored in practice. There is a high possibility that it will cause the spread of system failures. The only way to do this is to recreate and load the entire database.

However, in order to regenerate the entire database, it is necessary to stop providing the switch service, which is in operation.

It is an object of the present invention to provide a method for preventing interruption of service by recovering only a portion where damage occurs while operating a database in view of the conventional problems as described above.

In other words, the newly created relation data is read directly from the disk, and the database management system trap 2 (trap # 2) provided by the electronic switching system is used to write directly to the memory of the system. Writes to memory by obtaining the position information, size information, and characteristics of relation data to be loaded.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a diagram showing a relation file of an electronic switchboard disk and a relation file of a memory in the electronic switchboard processor used in the present invention.

2 is a flowchart illustrating a process of performing a data loading function according to the present invention.

Figure 3 is a view showing the interface between the operating system and the database management system and the application program in the electronic switching system to which the present invention is applied.

<Description of Symbols for Major Parts of Drawings>

1, 2: relation files 3, 4: application part

5: Operating System (OS) 6: Database Management System (DBMS)

In order to achieve the above object, in the present invention, when the database is found to be damaged during the database operation of the system, regenerating only the data of the damaged database to store on the disk;

Opening the data stored on the disk to collect identification number and header location information of a relation to be loaded from database header information;

Retrieving relation type information from the relation header to determine whether the relation to be loaded by collecting relation position information is a duplicate relation, and in the case of a duplicate relation, determining whether a duplicate relation duplication audit function is being performed;

When the duplicate audit function is not performed as a result of the determination, obtaining an address where actual relation data exists, the number of tuples, the size of the tuples, and the last tuple number of the data from the relation header information;

And multiplying the last tuple number by the tuple size to obtain a total data size, and reading the data to be loaded using the actual data address from the data size and the disk and transferring the data to a memory.

The above objects, features and advantages will become more apparent from the following detailed description taken in conjunction with the accompanying drawings. Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram showing a relation file of an all-electronic exchange disk and a relation file of a memory in the all-electronic exchange processor used in the present invention, wherein the memory or relational files 1 and 2 in the disk are normalized flat files or tables. Each column in a relation is called an attribute, and one row without a repeating group is called a tuple.

In this structure, the database stored in the electronic switch disk during the operation of the electronic switchboard is loaded into the relation file (1) address of the corresponding memory during initial setup of the system, and is updated or referred to by the database management system (DBMS) or an application. It is done.

At this time, the database loading function operates using the following information.

1) Database Header

Address of the relation header

2) relation header

-Relation identification number

-Relation type

Information about the data area in the processor

-Destination information according to relation type

Destination information according to relation location

-Number of attributes of this relation

The tuple size of the relation

-Address of relation existing in master address

-Relation lock information

-The maximum number of tuples given a master address

-Number of tuples actually used in master address

Number of the last tuple in the master address

FIG. 2 is a flowchart illustrating a process of performing a data loading function according to the present invention, in which data of a part determined to be damaged is already produced by using a database production tool, and the produced new data is stored in a database of a disc. Start with the saved state.

Since the process of producing such new data and storing it on the disk is already well known, detailed description thereof will be omitted.

The database file to be loaded from the disk is opened (S1), and the identification number and header position information of the relation to be loaded are collected from the database header information (S2).

After retrieving the relation type information from the relation header (S3), it collects the relation position information and checks whether the relation to be loaded is a duplicate relation (S4). When the function is being performed, the execution of the loading function ends (S5).

As a result of the determination, when the duplication audit function is not being performed, an address in which the actual relation data exists from the relation header information is obtained (S6).

In addition, the number of tuples, the size of the tuples and the last tuple number of the data are obtained from the relation header information (S7).

The total data size is obtained by multiplying the last tuple number by the tuple size (S8). After the size is obtained, the data to be loaded is read from the disk using the actual data address and the data size, and the data to be loaded is stored in a buffer of the memory, and then the identification number of the relation, which is loading information, the total size of the data to be loaded, and the actual data to be loaded are transferred to the memory. (S9).

3 is a diagram illustrating an interface between an operating system, a database management system, and an application program unit in an electronic switching system to which the present invention is applied. Each application program unit 3 and 4 includes a trap 1 and a trap. It interfaces with the operating system 5 and the database management system 6 through 2 (trap # 2), and the operating system 5 and the database management system 6 interface through the trap 0.

At this time, in the present invention, the position information of the relation in the memory is obtained by using the relation identification number transmitted in the memory transmission process (S9) of FIG. 2, and the actual data received is transmitted to the corresponding address through the trap # 2. Set to update.

For reference, the loading performing algorithm of the present invention performing the above operation is stored in the application program part and executed according to the request of the operator, and the operator recognizes that the damage occurs in the operating database. After the data is regenerated and stored in the disk, the algorithm of the present invention is operated and the loading position of the damaged database is found and loaded into the memory using the header information of the relation file stored in the disk.

Therefore, in the case of redundant relations, since the redundant relations are loaded into the memory of all processors in the electronic switchgear system, the inconsistency of the database is prevented, and the operator recovers by using the relation information in the system rather than directly by the operator. This prevents problems that can be caused by operator input errors.

As described in detail above, the present invention is to create a database file (relation file) to load the damaged part in the case of system failure due to database corruption by using the database production tool and to load only the relation into the system memory during system operation. This prevents service interruptions and maintains a stable system, recovers only damaged parts quickly, and in the case of redundant relations, it automatically loads into all subsystems where the relation is loaded, reducing the database inconsistency of the system. In addition to using the loading function, it ensures the reliability of the system and has the advantage of ensuring convenient and stable system operation.

In addition, a preferred embodiment of the present invention is disclosed for the purpose of illustration, those skilled in the art will be able to make various modifications, changes, additions, etc. within the spirit and scope of the present invention, such modifications and modifications belong to the following claims You should see.

Claims (1)

If database corruption is detected during system operation of the system, regenerating only the data of the damaged database and storing the data on a disk; Opening the data stored on the disk to collect identification number and header location information of a relation to be loaded from database header information; Retrieving relation type information from the relation header to determine whether the relation to be loaded by collecting relation position information is a duplicate relation, and in the case of a duplicate relation, determining whether a duplicate relation duplication audit function is being performed; When the duplicate audit function is not performed as a result of the determination, obtaining an address where actual relation data exists, the number of tuples, the size of the tuples, and the last tuple number of the data from the relation header information; And multiplying the last tuple number by the tuple size to obtain a total data size, and reading the data to be loaded using the actual data address from the data size and the disk and transferring the data to a memory. How to load relation data online.