JPS61844A - Data base system - Google Patents

Abstract

PURPOSE:To attain the quick and assured recovery of the meta-data by storing the log information on a meta-data base in a log file. CONSTITUTION:A meta-data file 1 is stored on an external recording medium in the form of the general data setting following the data setting of the data base. The contents of a meta-data base 2 are substantially equal to those of the file 1. The destruction of the base 2 can always be recovered by the file 1; while the destruction of the file 1 is backed up through the base 2. If both the base 2 and the file 1 are destructed, the meta-data is produced exclusive for the base 2 for recovery. Then the file 1 can be backed up. The data bases 3-1-3-n are set opposite to the log data for recovery of a log file 4. Thus the data bases 3-1-3-n can be recovered.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention provides a database system that enables data structure description data (meta data) for database recovery to be recovered under the management of a database management system.
It's about systems.

[Prior art and problems]

The log acquisition environment for recovery is shared among multiple databases. If a failure occurs, it is necessary to extract the log for a specific database. For this reason, each time a database is registered in the system, a new system-unique code is assigned to that database, and this code is used to manage logs and also perform recovery processing based on this code. Meta data is a correspondence table between databases and codes for this purpose. If a failure occurs in the meta data, it cannot be recovered, so methods such as duplicating the meta data are carried out, but simply duplicating the meta data does not require a high level of physical management. Otherwise, an incomprehensible state will remain in the meta data during processing, making it extremely difficult to understand the timing of recovery.

[Problem that the invention seeks to solve]

The present invention is based on the above considerations, and aims to provide a database system that allows meta data to be recovered reliably and at high speed.

[Means for solving problems]

Therefore, the database system of the present invention
A database, a meta data file that stores database meta data, a meta database that holds a copy of the meta data file, and a log file that holds recovery log data for the database and meta database. It is equipped with a file and a database management system having a transaction processing function,
In addition, when a request is made to update meta data, the above meta data
The present invention is characterized in that the data file is updated, this update is reflected in the meta data file, and log information of the meta database is stored in the log file.

[Embodiments of the invention]

Hereinafter, the present invention will be described in detail with reference to the drawings.

FIG. 1 is a diagram showing the relationship between data cents in the present invention. In FIG. 1, 1 is a meta data file, 2 is a meta database, 3-1 to 3-n are databases, and 4 is a log file.

The meta data file 1 is stored on an external recording medium as a general data center other than the data set of the database. The contents of meta database 2 are
The contents are substantially the same as data file 1.

If the meta database 2 is corrupted, it can be recovered from the meta data file 1 at any time, and conversely, if the meta data file 1 is corrupted, it can be banked up from the meta database 2. If both are corrupted, create meta data specifically for the meta database, recover meta data 2, and then back up meta data file 1. Each database 3-1. -.-3-n are associated with the recovery log data of the log file 4 by meta data, and thereby each database 3-1. -.-3-n becomes recoverable.

When there are hundreds of databases, databases are frequently registered and deleted. When registering or deleting a database, divide the meta data update process into multiple transaction processes, perform the 1st to Nth transaction processes in order, and update the meta data file 1 for each transaction process. Update, reflect this update on the meta database 2, and obtain recovery log data. Since the division of meta data update processing into transaction processing can be logically arbitrarily selected, logically meaningful breaks can be selected on the meta database 2. Logically meaningful means that the system can operate by copying the contents of meta database 2 to meta data file 1.

FIG. 2 shows an example of a configuration for carrying out the present invention. In FIG. 2, 5 indicates a database management system. Note that the same reference numerals as in FIG. 1 indicate the same parts. The database management system 5 has a transaction processing function. A database management system having a transaction processing function means that it is configured so that atomicity, permanence, and consistency of the database are established for each transaction process. For example, database 3-, n
When a request for registration comes, database 3
- Meta data related to n is meta data file 1
will be registered. As described above, this meta data registration process is performed by dividing into several serial transaction processes, and updates to the meta data file 1 are also reflected in the meta database 2 at the same time. In general, the meta data file 1 may have a physically more complex internal structure than the meta database 2, but in the meta database 2, anything below the physical level may be abstracted. Since the meta database entries in the meta data file 1 are permanent and unchanging from start to finish, the contents of the meta database entries can be reproduced at any time, and as long as this is available, transaction processing is possible. The database management system 4 has the following functions:
Recovery of the meta database 2 can be performed at any time using the meta data, so that a complete meta data file can be restored from this meta database 2.

〔Effect of the invention〕

As is clear from the above description, according to the present invention, the contents of meta data files and meta databases can always be recovered reliably and at high speed.

[Brief explanation of drawings]

FIG. 1 is a diagram showing the relationship of data centers in the present invention;
FIG. 2 is a diagram showing an example of a configuration for implementing the present invention. 1-meta data file, 2-meta database, 3-1 to 3-n-database, 4--log
File, 5-Database management system.

Claims (1)

[Claims] A database, a meta data file that stores database meta data, a meta database that holds a copy of the meta data file, and a log file that holds recovery log data for the database and meta database. It is equipped with a file and a database management system having a transaction processing function,
In addition, when a meta data update is requested, the above meta data
A database system characterized in that a data file is updated, this update is reflected in the meta data file, and log information of a meta database is stored in the log file.