KR100278024B1 - Apparatus and method for recreating a running database online - Google Patents

Abstract

The present invention relates to an apparatus and method for regenerating an online database online. When a reason for the reproduction of an operating database occurs in a system using a main memory resident database, the operating database is audited and backed up. The present invention relates to an apparatus and method for creating a new database online using a backed up database. The database regeneration process and the economical cost of saving the time and cost of applying the database to the exchange system are also provided. This simplifies and maximizes the operational state of the database, improving the efficiency of operations that eliminates the inefficiencies of manual input after applying data changes in the past due to the time difference between backup and application. It has the advantage.

Description

Apparatus and method for recreating a running database online

The present invention relates to an apparatus and a method for regenerating an online database online. The present invention relates to a system for using a main memory resident database. The present invention relates to an apparatus and method for auditing and backing up a running database, and then creating a new database online using the backed up database.

In general, exchange system software usually consists of applications and data. The data is part of the database, and the exchange system database is the main memory of the processor rather than the hard disk. The database management system (DBMS) that resides in and manages the database also resides in the processor's main memory.

If the database needs to upgrade the software version to add new service functions or to supplement operational problems, the database must be newly created because it involves changing / adding the relation due to the relation schema change besides the related application. The main reason for this is that the software version can not be processed quickly.

The exchange system database is produced offline at the database production host. As the reason for the database production occurs, it is necessary to preserve the current database to the maximum and produce a new database.

In the process, several problems occur. First, if the reason for making the database is simply a database schema change, it is unreasonable to back up the entire database and make it offline for some relations that need to be changed. The irrationality of re-entering the changes made to the database until the application of the database is applied, and then manually re-enter the database. Third, to apply the database when replacing the software version of the exchange system, the database backup, production, verification, and application It takes at least three to five days to complete the entire software version, compared to the amount of database changes, resulting in huge time and labor costs.

Therefore, there is a need for a method for creating a database by regenerating only relations according to schema changes while preserving data that does not need to be changed to the maximum.

In view of the above-described problems and requirements of the present invention, the present invention enables rapid database replacement by regenerating the database online in a system using a main memory resident database, thereby reducing the cost and time required for database creation. It aims to improve the efficiency of system operation by ensuring the maximum state of the database in operation.

In order to achieve the above object, the present invention performs an audit function for maximizing the inconsistency between the databases existing in the memory and the hard disk when necessary to rebuild the database of the exchange system using the main memory resident database You can back up the database, extract the relation object file from it, and change the relation object file as needed to create a new database.

1 is a schematic block diagram of an apparatus for regenerating a database online according to the present invention;

Figure 2 is a flow chart illustrating a database audit function processing process of the process of regenerating the operating database online in accordance with the present invention.

Figure 3 is a flow chart showing the processing of the function of changing the relation object file during the process of regenerating the operating database online in accordance with the present invention.

<Description of Symbols for Major Parts of Drawings>

10: changer 11: hard disk

12: system loader 13: memory

14: Database Audit Department

15: Database Backup / Dump section

20: database generator 21: database backup / dump

22: backup database extractor

23: relation object file changing unit 231: query parsing unit

232: query processing unit 24: database generation unit

30: database transmission path

In order to achieve the above object, in the present invention, a hard disk for storing a database to be used in the exchanger, a system loader for loading the database stored in the hard disk into the memory of the processor in the exchange, and the system loader unit Memory that resides the input database, database audit unit that performs the audit function for the database in operation when a reason for database re-production occurs, and database backup / dump unit that backs up and dumps the database stored in the hard disk. An exchanger provided;

A database generator for reproducing on-line the database currently running at the exchange;

And a database transmission path which is a transmission path for dumping a database between the switch and the database generator.

In addition, in order to achieve the above object, the present invention includes a first process of auditing the state of the database operating in the system using the main memory resident database;

A second step of backing up the audited hard disk database to a database generating means;

Extracting a relation object file from the backed up database;

A fourth step of changing the relation object file by inputting a query in accordance with relation schema information when there is a change to be made among the extracted relation object files;

And a fifth process of generating a new database from the relation object file extracted in the third process and the relation object file generated / modified in the fourth process.

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 block diagram schematically showing the configuration of an apparatus for producing a database online according to the present invention, comprising: a switch 10 resident in a main memory;

A database generator 20 for generating a database used in the exchange 10;

It is provided with a database transmission path 30 which is a transmission path for dumping the database between the switch 10 and the database generator 20.

The exchanger (10) includes a hard disk (11) for storing a database to be used;

A system loader (12) for loading a database stored in the hard disk (11) and loading it into the memory (13) of the processor in the exchanger (10);

A memory (13) for resident database input from the system loader (12);

A database audit unit 14 for performing an audit function on the database in operation when a reason for reproduction of the database occurs during operation of the switch 10;

And a database backup / dump unit 15 for backing up and dumping the database stored in the hard disk 11.

The database generator 20 includes a database backup / dump unit 21 for backing up and dumping a database generated by the database generator 20;

A backup database extraction unit 22 for extracting a relation object file from a new database inputted from the database backup / dump unit 12;

A relation object file change unit (23) for changing a relation object file for a relation whose schema is changed or newly added among the extracted relation object files;

And a database generator 24 for generating a new database using the extracted relation object file and the relation object file generated by the relation object file changer 23.

The relation object file changer 23 may include a query syntax parser 231 which analyzes a syntax of an operator when the operator inputs a query that is a data processing language of a relational database;

A query processing unit 232 for generating and changing a relation object file by changing relation schema and data according to the query analyzed by the query syntax analyzing unit 23 is provided.

Referring to the overall operation of the database online regeneration device configured as described above are as follows.

First, the process of applying the database to the exchange 10 is as follows. After the database is loaded onto the hard disk 11, the database in the hard disk 11 is loaded into the memory 13 of the processor by the system loader unit 12.

If a change in the relation data (log) occurs during operation, it is reflected in the database in the memory 13, the log is accumulated in a log buffer (not shown in the figure) periodically to the database in the hard disk 11 Is backed up.

In such a state, if a reason for database reproduction occurs, the switch 10 performs an audit function for the database in operation.

Such an audit function is performed through the upper database audit unit 14, and since a new database production is to backup and produce a database in the hard disk 11, it is necessary to audit the state of the hard disk 11 database.

This is because auditing before backing up the database in operation may occur because the inconsistency may occur in the database between the memory 13 and the hard disk 11 due to processor failure or incomplete execution of other applications.

After the audit is completed and the database inconsistency between the memory 13 and the hard disk 11 is resolved to the maximum, the database backup of the exchange 10 is carried out through the transmission path 30 between the exchange 10 and the database generator 20. The database of the hard disk 11 is backed up to the database generator 20 by the interaction of the database backup / dump unit 21 of the / dump unit 15 and the database generator 20.

In this case, the transmission path 30 between the switch 10 and the database generator 20 may be physically connected to an X.25 or local area network (LAN).

Subsequently, the database generator 20 extracts a relation object file from the backed up database through the backup database extracting unit 22.

The relation object file is changed by the relation object file changing unit 23 for a relation in which the schema is changed or newly added among the extracted relation object files.

In the changing method, when an operator inputs a query that is a data processing language of a relational database, the operator parses the query through the query parsing unit 231, and if the parse is parsed, changes the relation schema and data in the query processing unit 232 accordingly. Create and change the relation object file.

Subsequently, when the operation is completed, the database generator 24 creates a new database using the extracted relation object file and the relation object file generated by the relation object file changer 23.

The new database produced by this process is the hard disk 11 of the exchanger 10 by the interaction of the database dump unit 15, 21 between the exchanger 10 and the database generator 20 through the database transmission path 30. ), The system loader 12 loads the database into the processor's memory 13, thereby completing the database replacement.

2 is a flowchart illustrating a database audit function during a process of regenerating an operating database online according to the present invention.

Looking at the process using the drawings as follows.

In order to audit the state of the database in the hard disk 11, the memory 13 database is compared with the database header in the hard disk 11 (21).

In this case, the header information includes metadata such as relation directory, dictionary information, and a physical method of accessing the actual relation data, which are information necessary for extracting a database. The value generated at the time of manufacture contains invariant information that does not change even during operation. The invariant information is created and stored in a separate file when the database is initially stored on the hard disk 11.

When comparing the header, first, invariant information of the header of the memory 13 and the header of the hard disk 11 is compared (22). If the comparison result data is not the same, the invariant information in the hard disk 11 and the separate By comparing the invariant information stored in the file, it is checked whether the value of the hard disk 11 is valid (23).

If the comparison result is different, the corresponding header information of the memory 13 and the hard disk 11 is changed to the value in the fixed information table (24), and then the header of the memory 13 and the hard disk 11 are next. The variable information in the header is audited (25).

If the comparison result is different, it checks whether the relation data (number of tuples) of the memory matches the header information of the memory (26), and if so, the memory header information and the actual data are copied to the corresponding area of the hard disk (28).

If they do not match, recovery cannot be performed (27).

On the other hand, if the comparison result 25 of the variable information is the same, each time the data and memory data change of the hard disk 11 occurs, the memory 13 tuple data and the hard to check whether it is maintained consistently and consistently The disk 11 tuple data is compared (29).

If the comparison result is different, the data mismatch between the memory 13 and the hard disk 11 has occurred, and thus the data of the memory 13 is copied to the corresponding area of the hard disk 11 (30).

If the comparison result 29 is the same, the database of the hard disk 11 reflects the memory database well since a normal and successful backup is performed every time a memory 13 database change occurs. It can be seen that.

3 is a flowchart illustrating a process of a function of changing a relation object file during a process of regenerating a database in operation according to the present invention online.

The relation object file is composed of a header portion describing attribute information and tuple information of the relation and an actual relation tuple data area.

In the creation and modification of the relation object file, if a query suitable for each purpose is inputted, the query parser 23 analyzes the input query and executes the corresponding query. At this time, the header and data area of the relation are changed as necessary. do.

When a new relation is to be added according to the above change (31), a relation header is generated with the schema information inputted by inputting a dictionary, tuple number, storage type, relation type information, etc., which are attribute information of the relation, in the form of a query. (32).

Subsequently, an area for storing relation tuple information is allocated according to the number of tuples, and tuple data is generated with default values of each property (33).

On the other hand, processing when a schema such as adding or changing an attribute of an existing relation is changed is as follows (34).

When the relation attribute is added (35), the relation header is changed according to the input attribute information (36), and the region for the attribute added to the relation data region is secured and filled with the initial value (37).

In the case of the change of the existing attribute (38), the changed attribute information is reflected in the relation header (39) and the attribute size is increased when the range of the domain data type of the attribute is increased to store the larger value. Since it may be increased, the area to store the value is increased according to the increased attribute size as needed (40).

In addition, when the number of tuples of a relation is increased or decreased (41), the changed tuple information is reflected in the relation header (42), and the area for tuple data is adjusted by the number of increased or decreased tuples, and if the number of tuples is increased, As much as the tuple is filled with the initial value (43).

On the other hand, processing in the case of changing relation tuple data instead of changing relation schema is as follows (44).

When the tuple is inserted (46), the tuple information is inserted by inserting tuple information of the relation through the tuple insertion query (47).

In addition, when the attribute value needs to be changed (48), the tuple matching the condition of the relation is found through the change query word, and the desired attribute value is changed (49).

When the tuple needs to be deleted (50), the tuple matching the condition is found in the relation through the tuple deletion query word and the tuple is deleted (51).

Finally, if the above is not the case, check the relation data once again (45).

As described in detail above, the present invention has an economical advantage in reducing the time and cost of database creation as well as the application of the database to the exchange system, and the database reproduction process is simplified and the state of the database in operation is guaranteed to the maximum. The method has the advantage of improving the efficiency of the operation, eliminating the inefficiency of manually entering data changes in the past and applying them manually after the time difference between backup and application.

Therefore, according to the present invention, the production and application of the database, which has been an obstacle to the application of the exchange system software, is made online, so that the application of the exchange system software version can be performed quickly to meet various needs of subscribers according to the development of information and communication technology. This has the advantage of forming a foundation for providing new services quickly.

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 will have to look.

Claims (16)

A hard disk storing a database for use in the exchanger, a system loader unit for loading the database stored in the hard disk into the memory of the processor in the exchanger, a memory for residing the database input by the system loader unit, and the exchanger in operation An exchanger having a database auditing unit for performing an audit function on the database in operation, and a database backup / dump unit for backing up and dumping a database stored in the hard disk when a database re-production reason occurs; A database generator for reproducing on-line the database currently running at the exchange; And a database transmission path which is a transmission path for dumping a database between the switch and the database generator. The method of claim 1, The database generator includes a database backup / dump unit for backing up and dumping a database generated by the database generator; A backup database extraction unit for extracting a relation object file from a new database inputted from the database backup / dump unit; A relation object file change unit for changing a relation object file with respect to a relation whose schema is changed or newly added among the extracted relation object files; And a database generator for generating a new database by using the extracted relation object file and the relation object file generated by the relation object file changer. The method of claim 2, The relation object file change unit may include: a query syntax parser configured to analyze a syntax of an operator when the operator inputs a query that is a data processing language of a relational database; And a query processing unit configured to generate and change a relation object file by changing relation schema and data according to the query analyzed by the query parsing unit. A first step of auditing a database state in operation in a system using a main memory resident database; A second step of backing up the audited hard disk database to a database generating means; Extracting a relation object file from the backed up database; A fourth step of changing the relation object file by inputting a query in accordance with relation schema information when there is a change to be made among the extracted relation object files; And a fifth process of generating a new database from the relation object file extracted in the third process and the relation object file created / modified in the fourth process. The method of claim 4, wherein The first process includes a first step of comparing invariant information of a memory header and a hard disk header; A second step of comparing invariant information in the hard disk with fixed information table data for the invariant information when the comparison result data is not the same; A third step of performing audit on variable information of a memory header and a hard disk header after changing corresponding header information of a memory and a hard disk to a value stored in a fixed information table if the comparison result value is different; And if the audit results are different or the same, performing a corresponding process, and then copying the memory data to the corresponding area of the hard disk. The method of claim 5, A first step of checking whether relation data of the memory matches header information of the memory if the variable information comparison result is different in the fourth step; And a second step of copying the memory header information and the actual data to the corresponding area of the hard disk if the comparison result is matched. The method of claim 5, A first step of checking whether the memory tuple data and the hard disk tuple data match if the variable information comparison result is the same in the fourth step; And if the comparison result is different, performing a second step of copying the memory tuple data to the corresponding area of the hard disk. The method of claim 5, The fixed information table is a table generated by storing the data for fixed information on the hard disk and storing the data in an independent table at the time of storing the database on the initial hard disk. How to. The method of claim 4, wherein The changing of the relation object file in the fourth process may include: a first step of executing the query by analyzing the input query through a query syntax parser when a query suitable for generation and modification of the relation object file is input; When a new relation needs to be added when a change is made according to the query analysis, a relation header is generated by inputting a dictionary, tuple number, storage type, relation type information, etc., which are attribute information of the relation, in a query form and using the input schema information. Two steps; And generating a tuple data with an initial value of each attribute by allocating an area for storing relation tuple information according to the number of tuples. The method of claim 9, A first step of inserting attribute information input in the relation header when the attribute of the relation is added when the change according to the query analysis in the first stage is a relation schema modification; And a second step of storing an initial value by securing an area for an attribute added to the relation data area. The method of claim 10, A first step of reflecting the changed property information in the relation header when the existing property is changed during the relation schema change in the first step; And regenerating the operating database online, comprising: a second step of expanding an area to store a value according to the increased attribute size as needed when the size of the attribute domain data type is increased according to the changed reason. How to. The method according to claim 10, A first step of reflecting the changed tuple information in the relation header when the number of tuples of the relation is increased or decreased during the relation schema change in the first stage; Adjusting a region for tuple data by the increased or decreased number of tuples; And a third step of storing the initial value of the tuple by the increased number of tuples when the number of tuples is increased. The method of claim 9, If the tuple is inserted when the change according to the query analysis in the first step is a relation data change, the tuple information is input by inserting tuple information of the relation through a tuple insertion query. To regenerate. The method of claim 13, If the attribute value needs to be changed when the change according to the query analysis in the first step is a relation data change, the desired attribute value is changed by finding a tuple that matches the condition of the relation through the change query word. How to recreate a database online. The method of claim 13, If the tuple is to be deleted when the change according to the query analysis in the first step is a relation data change, the tuple deletion query word searches for a tuple matching a condition in the relation and deletes the tuple online. To regenerate. The method of claim 4, wherein The relation object file includes attribute information of the relation; A header portion describing tuple information; A method for regenerating a live database online comprising a real relation tuple data area.