Types of Failure

An unfortunate aspect of every database system is the possibility of a system or hardware failure. The most common types of failure are as follows.

• Media failure, when one or more of the disk drives holding a database fails and you are faced with a complete loss of data unless you can restore and recover from a backup of the database.

• Application errors, when an application makes a large number of invalid modifications to data and the best way to deal with the problem may be to restore and recover the database from a backup to a specific date and time before the modifications were made.

• Software failure, when the operating system, device drivers, environmental software, or JADE itself has a problem that leads to file system or database corruption. For example, the File Allocation Table (FAT) file system is susceptible to various forms of corruption.

• Permanent loss of a server, when a server is disabled permanently or a site is lost due to a natural disaster and you may need to activate a warm standby server or restore and recover from a backup of the database on another server. For details about standby servers, see "Roll-Forward Recovery of a Standby Database", later in this chapter.