This is a "made up" term just to catch you out! Locking protects against loss of information, and so this is not the correct answer. If a transaction needs access to an attribute which is locked, then that transaction is suspended until the attribute is available. However, if that transaction has already locked something which other transactions need, then it is possible to enter a state where the transactions of a system are all suspended and the attributes which they are waiting on will never become available. This is called deadlock and can result when locking is introduced into a system. Locking enforses serialisability, and thus enforces integrity. Look again at the other options. What is the biggest problem which results from the introduction of locking? Which one of the following problems can occur due to introducing locks in a concurrent transaction scenario?
|
|