JP2016212480A - Database system, database backup system, and database backup program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To shorten cold back time of databases, and reduce a capacity of backup retreat areas.SOLUTION: A database system according to the present invention comprises control means that controls backup processing of a file data group of a first database in a second database. The control means includes: an initial size setting unit that specifies a minimum data storage area for each file category of the second database, and sets an initial size of the second database; a cold backup processing unit that executes a cold backup of the file data group existing in the first database to the second database; and a storage area size expansion unit that expands the data storage area of the file data group in the second database after execution of the cold backup.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a database system, a database backup system, and a database backup program. For example, the present invention can be applied to a system that performs database backup processing by a database cold backup method.

  FIG. 2 is an explanatory diagram for explaining cold backup of a conventional database.

  In general, a cold database backup is a backup of the database configuration, and the database file system (database) 2-1 is kept in a state where the database service, that is, the RDBMS (Relational database management system) of the database server is stopped. This means copying a set of physical files to be configured and saving the set of physical files to the cold backup save file system 2-2.

  Here, as illustrated in FIG. 3, the physical file set includes a control file 21 for the RDBMS to control the database, a parameter file 22 that records the configuration of the database, a data file 23 that stores the database data, There are a log file 24 that records a log at the time of data update, a rollback file 25 that records information for rolling back the update, a temporary work file 26, and the like.

  In contrast to cold backup, the RDBMS 11 has a function of dumping database data, that is, a data group stored in a data file to a file. For example, in the case of an ORACLE (registered trademark) database, the export function corresponds to this data dump function.

  Data dumping can be performed without stopping the database operation, and the time required for file output is shorter than that for cold backup.

  Database recovery from backup can be completely recovered by these cold backups and exported dump files.

  Normally, as shown in FIG. 4, the cold backup of the database is acquired before the full-scale operation of the database (S501), and after the service is started, dumping is periodically performed (S502 to S504). Periodic dumping is effective at the time of failure recovery because the data updated so far can be backed up.

JP 2011-186902 A

  However, in the case of a large-scale database, for example, a conventional database backup system constructs a database after estimating the amount of stored data when designing the database. Therefore, when acquiring a cold backup of the database, the time required for the backup may be very long in proportion to the physical file size.

  In particular, a data file or a log file may be composed of a plurality of files, which is one of the factors that increase the time for cold backup.

  Therefore, there is a need for a database system, a database backup system, and a database backup program that can reduce the cold-back time of the database by reducing the initial size of the database to be constructed and also reduce the capacity of the backup save location. ing.

  In order to solve such a problem, a database system according to a first aspect of the present invention includes a first database, a second database that is a backup destination of a file data group physically existing in the first database, In a database system comprising control means for controlling backup processing of a file data group of one database to a second database, the control means performs (1) for each file type of the second database that backs up the file data group. An initial size setting unit that specifies the minimum data storage area size and sets the initial size of the second database; and (2) a cold that cold-backups the file data group existing in the first database to the second database. Backup processing unit, (3) After cold backup And having a storage area size expanding portion for expanding the data storage area size of the file data group in the second database.

  A database system according to a second aspect of the present invention provides a database backup system that performs backup processing on a second database that is a backup destination of a file data group physically existing in the first database. Initial size setting means for setting the initial size of the second database by designating the minimum data storage area size for each file type of the second database to be backed up, and (2) file data existing in the first database A cold backup processing means for performing cold backup of the group to the second database; and (3) storage area size expanding means for extending the data storage area size of the file data group of the second database after the cold backup is performed. Features.

  A database backup program according to a third aspect of the present invention is a database backup program for performing backup processing on a second database, which is a backup destination of a file data group physically existing in a first database. An initial size setting means for setting the initial size of the second database by designating the minimum data storage area size for each file type of the second database for backing up the data group; and (2) existing in the first database. Cold backup processing means for performing cold backup of the file data group to be stored in the second database, and (3) storage area size extending means for extending the data storage area size of the file data group of the second database after the cold backup is performed. Preparation And wherein the Rukoto.

  According to the present invention, by minimizing the initial size of the database to be constructed, it is possible to shorten the database cold-back time and to reduce the capacity of the backup save location.

It is a block diagram which shows the whole structure of the database system which concerns on 1st Embodiment. It is explanatory drawing explaining the cold backup of the conventional database. It is explanatory drawing explaining the file group of a set of physical files. It is explanatory drawing explaining the dump by cold backup and export of a database. It is explanatory drawing explaining the comparison of the database construction size which concerns on 1st Embodiment. It is a flowchart which shows the process sequence until it implements the database backup in DB system which concerns on 1st Embodiment. It is a figure which shows the operation database size of DB file system which concerns on 1st Embodiment. It is explanatory drawing explaining the example of the database construction command which concerns on 1st Embodiment. It is explanatory drawing explaining the file size of DB file system constructed | assembled by the cold backup which concerns on 1st Embodiment. It is explanatory drawing explaining the example of the cold backup command which concerns on 1st Embodiment. It is explanatory drawing explaining the example of the file compression command which concerns on the cold backup which concerns on 1st Embodiment. It is explanatory drawing explaining the example of the size expansion command of the database after the cold backup which concerns on 1st Embodiment. It is explanatory drawing explaining the example of the SQL command which deletes the log file group after the cold backup which concerns on 1st Embodiment. It is explanatory drawing explaining the example of the OS command which deletes the log file after the cold backup which concerns on 1st Embodiment.

(A) First Embodiment Hereinafter, a first embodiment of a database system, a database backup system, and a database backup program according to the present invention will be described in detail with reference to the drawings.

(A-1) Configuration of First Embodiment FIG. 1 is a block diagram showing an overall configuration of a database system (hereinafter also referred to as a DB system) 10 according to the first embodiment. In FIG. 1, a DB system 10 includes a database server (hereinafter also referred to as a DB server) 1 and a database file system (hereinafter also referred to as a DB file system) 2 (2-1 and 2-2). The number of DB file systems 2 in the DB system 10 is not particularly limited.

  The DB server 1 is connected to DB file systems 2-1 and 2-2 for storing database data. Further, the DB server 1 has a server process 11 that provides a service using a database stored in the DB file system 2 to a DB client (not shown) (executes transaction processing).

  The server process 11 can be constructed by using database middleware such as RDBMS, for example. In this embodiment, tables (tables) for managing databases are stored in the DB file systems 2-1 and 2-2, and the server process 11 responds to a request from a DB client (not shown). Table) data is updated.

  The server process 11 backs up a set of physical files stored in the DB file system 2-1 to the DB file system 2-2 by a cold back method. The DB file system 2-1 is an operational database, and the DB file system 2-2 is an initial database.

  Here, when the server process 11 backs up a set of physical files to the DB file system 2-2 as an initial database by a cold backup method, the server process 11 constructs the initial database (DB file system 2-2) in a small size.

  In addition, the server process 11 expands to the estimated size at the time of designing the database (that is, the database size of the operating conditions) after performing the cold backup. Thus, the backup work time can be shortened and the backup file capacity can be reduced.

  Regarding the point that the DB file system 2-2, which is the initial database, is constructed in a small size, the data storage area when the database is defined by the server process 11 using the database creation command during database construction is reduced to the minimum. This can be realized by specifying the specified size.

  FIG. 5 is an explanatory diagram illustrating comparison of database construction sizes according to the first embodiment. For example, as shown in FIG. 5, in the DB file system 2-1 to be constructed, even when the size of the data file 23A requires 10 Gbytes (this calculated value at the time of design estimation), the database usage rate before the start of full-scale operation is It should be extremely small (or it can be said that the number of data is 0 immediately after the database construction). Therefore, it can be said that the data storage area size of the data file 23B of the DB file system 2-2 of the initial database can be created with a size of 256 Mbytes or less.

  Also, for the log files 24A and 24B, the rollback files 25A and 25B, and the temporary work files 26A and 26B, similarly to the data files 23A and 23B, the data storage area size of the initial database of the DB file system 2-2 is made small. can do.

  Furthermore, after cold backup is performed, file compression processing is generally performed to reduce the size of the backup file. In this case as well, the size of the initial database can be reduced, and file compression with almost no registered physical file data can increase the compression efficiency and reduce the file size for backup storage as much as possible. Become.

  For example, in the case of the ORACLE (registered trademark) database, the control files 21A and 21B and the parameter files 22A and 22B are not specified in size when the database is created, and the file size itself is small. It can be said that the ratio included in the entire time is small.

  As shown in FIG. 5, in the cold backup of the DB file system 2-1, after the database is constructed, the DB file system 2-2 acquires a set of physical files before full service operation. After acquiring a set of physical files using the cold backup method, the file size is expanded to meet the requirements of the DB system. This size expansion can also be performed by executing a command provided by the service process 11.

  According to the procedure of the cold backup method of the first embodiment, the time required for implementing the cold backup method of the database can be shortened as much as possible, and the file size for backup storage can be made as small as possible.

(A-2) Operation of First Embodiment Next, the database backup processing operation according to the first embodiment will be described in detail with reference to the drawings.

  FIG. 6 is a flowchart illustrating a processing procedure until database backup is performed in the DB system 10 according to the first embodiment.

  First, it is assumed that the DB file system 2-1 (operation database) that satisfies the size requirements necessary for the DB system 10 has the operation database size shown in FIG. 7. At this time, the service process 11 constructs a DB file system 2-2 as an initial database by using the SQL command shown in FIG. 8 (S101). This process is also called a database construction stage.

  For example, in FIG. 7, it is assumed that the size of the data file 23A is 10 Gbytes, the size of the log file 24A is 10 Gbytes, the size of the rollback file 25A is 10 Gbytes, and the size of the temporary work file 26A is 10 Gbytes.

  The SQL command shown in FIG. 8 is a database construction command of the DB file system 2-2. The data file size is specified (size: 256 Mbyte), the log file size is specified (size: 256 Mbyte), and the rollback file size is specified ( Size: 256 Mbytes) and designation of a temporary work file (size: 256 Mbytes). As described above, the service process 11 constructs an initial size database by designating the initial size of the data file 23B, the log file size 24B, the rollback file 25B, the temporary work file 25B, and the like.

  In addition to this command, there is something that should be executed before actually building a complete database, but it is omitted here.

  Next, when the database of the initial size is constructed, the files shown in FIG. 9 that are the targets of the cold backup are generated, and these files are cold backed up.

  That is, the service process 11 executes a command for copying a file after stopping the database (S102). Thereby, cold backup is performed (S103). For example, when the OS environment is Linux (registered trademark), the cp command may be executed. If file compression is to be performed, a gzip command may be executed.

  The DB file system 2-2 copies the data file as a backup file to the “/ backup” directory, and in order to compress the backup file, the OS command “cp / var / test / data / system01” shown in FIG. .Dbf / backup / ”,“ gzip / backup / test / system01.dbf ”.

  The db file system 2-2 completes cold backup by repeating copying and compression in order with respect to a file to be subjected to cold backup.

  At this time, the reason for stopping the database is that access to the database occurs while the file is being copied, and the contents of the file are rewritten. This is to prevent inconsistencies.

  After completion of a series of cold backups, the database of the DB file system 2-2 is activated (S104), and the service process 11 expands the DB file system 2-2 to a size for the operation database (S105).

  Expansion of the database size of the DB file system 2-2 is executed by the service process 11 instructing the SQL command shown in FIG. As shown in FIG. 11, the SQL command of the service process 11 extends the data storage area size of the data file 23B, the rollback file 24B, and the temporary work file 26B to 10 Gbytes, for example.

  Note that the size extension of the log file 24B is different from the above command. Specifically, this is a procedure for adding a log file different from the log file created at the time of initial database construction and deleting the log file at the time of initial database construction.

  As an example, first, in order to add a new log file group, the service process 11 performs the SQL command “ALTER DATABASE ADD LOGFILE GROUP 2 ′ / var / test / redo / redoA0201.log ′, ′ / u01 shown in FIG. /Orasid/redo/redoA0202.log ') SIZE 10240M; ".

  Next, in order to delete the log file group 1, the service process 11 executes the SQL command “ALTER DATABASE DROP LOGFILE GROUP1;” shown in FIG.

  Finally, in order to delete the log file, the service process 11 uses the OS commands “rm / var / test / redo / redoA0101.log” and “rm / var / test / redo / redoA0101.log” shown in FIG. Run.

  Note that the actual log file group addition or deletion requires status confirmation, log switching work, etc., but since this is a part not related to the processing of the first embodiment, details here The detailed explanation is omitted.

  Here, since the log file (GROUP2) is generated as a new file, it will be different from the one that was cold-backed. However, if the database is restored using the cold-backup file, Means that the file may be regenerated by the above procedure after recovery.

  If the log file is created from the beginning with 10 Gbytes, it will be necessary to copy and copy back the 10 Gbyte log files both during cold backup and during recovery. Only the creation time of a new log file (10 Gbyte GROUP 2 file) is sufficient.

  The procedure for extending the size of the operation database from the initial database construction to cold backup was described above.

  This procedure completes the database according to the original requirements, and the actual operation, that is, the service can be started.

(A-3) Effect of First Embodiment As described above, according to the first example, the database backup time can be reduced by setting the size at the time of database construction as the minimum unit. Can do.

(B) Other Embodiments In the first embodiment, the method of the present invention is implemented in the server process 11 constituting the RDBMS. For example, the present method is implemented as a process different from the RDBMS of commercial middleware. However, commands provided by the RDBMS may be controlled.

  For example, in the case of an ORACLE database, a set of commands necessary for database construction according to this method is a function that has already been implemented, and the execution conditions, execution triggers, and parameters of the command are the points of this method.

  DESCRIPTION OF SYMBOLS 10 ... Database system, 1 ... Database server, 11 ... Service process, 2-1 ... Database file system, 2-2 ... Database system.

Claims (4)

A first database;
A second database that is a backup destination of the file data group physically existing in the first database;
A database system comprising: control means for controlling backup processing of the file data group of the first database to the second database;
The control means is
An initial size setting unit that specifies a minimum data storage area size for each file type of the second database that backs up the file data group, and sets an initial size of the second database;
A cold backup processing unit that cold-backups the file data group existing in the first database to the second database;
A database system comprising: a storage area size expansion unit that expands a data storage area size of the file data group of the second database after performing a cold backup.   The storage area size expansion unit extends the storage area size of each file type in the second database so as to match the storage area size of the file type in the first database as a backup source. The database system according to claim 1. In a database backup system for performing backup processing on a second database that is a backup destination of a file data group physically existing in the first database,
An initial size setting means for specifying a minimum data storage area size for each file type of the second database for backing up the file data group, and for setting an initial size of the second database;
Cold backup processing means for cold backing up the file data group existing in the first database to the second database;
A database backup system, comprising: storage area size expansion means for expanding a data storage area size of the file data group of the second database after performing cold backup. In a database backup program for performing backup processing on a second database that is a backup destination of a file data group physically existing in the first database,
An initial size setting means for setting a minimum data storage area size for each file type of the second database for backing up the file data group and setting an initial size of the second database;
Cold backup processing means for cold backing up the file data group existing in the first database to the second database;
A database backup program comprising storage area size expansion means for expanding a data storage area size of the file data group of the second database after performing a cold backup.

Images