Retrieving Database Concepts

4:11 PM
Retrieving Database Concepts -

The process of solving any kind of database errors, quickly and without loss of data and maintain high available database is called the main elements of recovery.The database database recovery is the most recent of backup.If database is kept up the database efficiently, then the database recovery is very simple process forward.

The components and database technologies is to set to reduce to a minimum or no loss of data and maintain the database to open, includes

* Checkpoints

* control file

* redolog online file

* archived redo log files

* Flash Recovery Area

Recovery Technologies

* Flash database technology back

* Open the database in Archive log mode

* Implement highly available database features such as RAC, DataGuard, vapors, etc.

Components RDB

Checkpoints: - Checkpoint (CKPT) in the Oracle database is a mandatory background process. This process works in accordance with the database writer process (DBWn) to handle the amount of time required for the recovery instance. Checkpoint in Oracle occurs when there is the manual or automatic log switch. FIRST_START_MTTR_TARGET parameter can be adjusted to control the result of checkpoints to improve sample recovery

control file -. control file is a repository of metadata for the database. It keeps all the information on the database structure. control file is very critical for the operation of the database, in order to be a DBA, you should keep at least two copies of the control file (Oracle recommend three) and save them in a different disk for the same faults disk file control.

Multiplexing control files

* Through init.ora

* Through spfile

Multiplexing control files in the CONTROL_FILE though init.ora initialization parameter file

CONTROL_FILES =('/u01/app/oracle/oradata/prod/ctlorcl01.ctl','/u02/app/oracle/oradata/prod/ctlorcl02.ctl','/u03/app/oracle/oradata/prod/ctlorcl03.ctl')

Storing control files in multiple position protect the control file from the single point of failure. In multiplexing control configuration files, Oracle updates all the control files at a time, but only use one, first listed in CONTROL_FILES parameter.When a single control is corrupt, then the DBA can simply copy a good one for that position or any other place and make the change in CONTROL_FILE parameter in the init.ora file, solve the problems.

Through SPFILE

We can dynamically set or modify the control file location using SPFILE concept using the following command.

SQL> ALTER SYSTEM SET CONTROL_FILES =

'/ u01 / app / oracle / oradata / prod / ctlorcl01.ctl',

'/ u02 / app / oracle / oradata / prod / ctlorcl02.ctl',

'/ U03 / app / oracle / oradata / prod / ctlorcl03.ctl') scope = spfile ;.

The parameter will affect when you restart the Oracle

instance

Multiplexing redo log files

information in online redolog files are very important for the recovery of the database. To preserve these files against any kind of damage you can configure multiple network file copies for redolog different location on the disc. All copies of redolog multiplex files are the same size and in a group. Each redolog file in a group is called members. background process LGWR writes redo information to all members of the same group, it is a best practice to protect the redolog files from the single point of failure. When multiplexing redolog files, group members should keep in different disks so that a disk failure will not affect the normal Normal operation.For database operation database at least two redo log groups is required.

Creating a new log file group

We can create new redolog groups using the following command.

SQL> ALTER DATABASE ADD LOGFILE

GROUP 4 ( '/u01/app/oracle/oradata/prod/redo0401.log',

'/ u01 / app / oracle / oradata / prod / redo0402.log'

'/ u01 / app / oracle / oradata / prod / redo0403.log') SIZE 10M ;.

We can add a new member to an existing group using the following command

SQL> ALTER DATABASE ADD LOGFILE MEMBER

'/ u01 / app / oracle / oradata / prod / redo0401.log' to group 3;

members renaming log

followed by steps to follow to rename the log members

1. Shutdown database.

2. Copy / rename the redolog files to the new location with the operating system command

3. based startup data with the STARTUP MOUNT command

4. Run SQL > ALTER databse rename the 'OldFileName' TO 'newfilename' file

5. ALTER DATABASE OPEN

Dropping redolog groups

SQL> ALTER databse DROP LOGFILE gROUP 3

Previous
Next Post »
0 Komentar