Oracle Redo Log Files
Posted: Tue Jan 28, 2025 3:23 am
Fig.6: Logging in the redo log buffer. © Marek Adar
Fig.6: Logging in the redo log buffer. © Marek Adar
In order to save the changes to the database buffer cache in advance, the contents of the redo log buffer are written to the redo log files. A database must always have at least two of these files, but usually has more for performance reasons.
These files are written to cyclically. This means: if the first file is full, the system switches to the second file and continues writing there; if this file is also full, the system switches back to the first file and continues writing. The redo log files therefore contain information about the data record changes in the blocks and are used to restore the changed blocks of the database in the event of an instance crash.
In earlier versions of Oracle, the changed blocks of the database buffer cache were written to a new redo log file in the database after each switch. In afghanistan telegram screening newer versions, the write-back is carried out according to rules that are not specified here, but at the latest when the switch is made back from the last file to the first.
Since the redo log files are permanently written, they should not be stored on the same disk subsystem as the operating system or the database, but on a separate disk subsystem. This prevents access conflicts when writing the data, which could affect the transaction speed.
In addition, it is necessary to mirror these files via Oracle, since a failure of the active redo log file would result in a complex recovery process, because all data changes written to the active file have not yet been transferred to the database.
Fig.6: Logging in the redo log buffer. © Marek Adar
In order to save the changes to the database buffer cache in advance, the contents of the redo log buffer are written to the redo log files. A database must always have at least two of these files, but usually has more for performance reasons.
These files are written to cyclically. This means: if the first file is full, the system switches to the second file and continues writing there; if this file is also full, the system switches back to the first file and continues writing. The redo log files therefore contain information about the data record changes in the blocks and are used to restore the changed blocks of the database in the event of an instance crash.
In earlier versions of Oracle, the changed blocks of the database buffer cache were written to a new redo log file in the database after each switch. In afghanistan telegram screening newer versions, the write-back is carried out according to rules that are not specified here, but at the latest when the switch is made back from the last file to the first.
Since the redo log files are permanently written, they should not be stored on the same disk subsystem as the operating system or the database, but on a separate disk subsystem. This prevents access conflicts when writing the data, which could affect the transaction speed.
In addition, it is necessary to mirror these files via Oracle, since a failure of the active redo log file would result in a complex recovery process, because all data changes written to the active file have not yet been transferred to the database.