Database can fail due to some reasons like media
failure, operating system failure, accidental damage, and intentional
damage to the database. When database stops working due to any above
reasons, it’s the responsibility of Database Administrator to recover
the database using its backup files and restoration of database into the
state it was in, before its failure. Normally full backup are done after
one week but incremental backup are done on daily basis.
In these days, almost every database management
system gives a number of methods to restore the failure database using
its backup. It depends upon the size of backup file, when the size of
database is large, recovery process may take a lot of time. There are
two major types of complete recovery the database.
Backward Recovery
In backward Recovery the database is restored into
the previous state and the unwanted changes are eliminated. It means
that all changes made by various transactions are undone. For example,
it will be equivalent to the state that was fifteen days before the
failure. This sort of recovery is made when up to the time backup is not
available or changes of previous period are needed to be undone.
Forward Recovery
It is also called Roll-forward. Changes of some failed transactions are
applied to the database in order to roll it forward. Thus database
becomes updated with all the changes confirmed.