The assumption that the hardware errors bring the system to a halt is called as fail-stop assumption. Well-designed systems have numerous internal checks that bring a system to a halt whenever there is an error. Hence, the fail stop assumption is a reasonable one.

Logical errors and system errors are the two types of errors that can generally cause transaction failures.

If a system has entered and undesirable state due to which it is unable to continue with normal execution, it is called as a system error. The transaction however can be re-executed at a later time.

The database is partitioned into fixed length storage units called as blocks. Blocks are the units of data transfer and may contain several data items.

If a failure has occurred in the midst of a transfer, it is called as partial failure. After a partial failure, the destination has incorrect information.

Volatile storage, non-volatile storage and stable storage are three different classifications of storage. Volatile storage is the least secure whereas stable storage is the most secure among them.

The blocks residing on the disk are referred to as physical blocks. The blocks residing temporarily in the main memory are referred to as buffer blocks.

If the transaction can no longer continue with its normal execution because of some internal condition, it is called as a logical error such as bad input, data not found, overflow or resource limit exceeded.

If there is a hardware malfunction or a bug in the database that causes the loss of content of volatile storage, it is called as a system crash. The data in the non-volatile storage remains intact and it is not corrupted.

In case of total failure, the failure occurs sufficiently early during the transfer that the destination block remains intact. Thus there is no corruption of data at the destination.