Databases Reference
In-Depth Information
concept is that such systems should certainly be expected to handle these issues
frequently for all the data stored in the system's databases.
Computer security has become a very broad topic with many facets and
concerns. These include protecting the physical hardware environment, defending
against hacker attacks, encrypting data transmitted over networks, educating
employees on the importance of protecting the company's data, and many more. All
computer security exposures potentially affect a company's data. Some exposures
represent direct threats to data while others are more indirect. For example, the theft
of transmitted data is a direct threat to data while a computer virus, depending on
its nature, may corrupt programs and systems in such a way that the data is affected
on an incidental or delayed basis. The types of direct threats to data include outright
theft of the data, unauthorized exposure of the data, malicious corruption of the
data, unauthorized updates of the data, and loss of the data. Protecting a company's
data assets has become a responsibility that is shared by its operating systems,
special security utility software, and its database management systems. All database
management systems incorporate features that are designed to help protect the data
in their databases.
Data can be lost or corrupted in any of a variety of ways, not just from the
data security exposures just mentioned. Entire files, portions of databases, or entire
databases can be lost when a disk drive suffers a massive accidental or deliberate
failure. At the extreme, all of a company's data can be lost to a disaster such as
a fire, a hurricane, or an earthquake. Hackers, computer viruses, or even poorly
written application programs can corrupt from a few to all of the records of a file
or database. Even an unintentional error in entering data into a single record can
be propagated to other records that use its values as input into the creation of their
values. Clearly, every company (and even every PC user!) must have more than
one copy of every data file and database. Furthermore, some of the copies must be
kept in different buildings, or even different cities, to prevent a catastrophe from
destroying all copies of the data. The process of using this duplicate data, plus
other data, special software, and even specially designed disk devices to recover
lost or corrupted data is known as ''backup and recovery.'' As a key issue in data
management, backup and recovery must be considered and incorporated within the
database management system environment.
In today's multi-user environments, it is quite common for two or more users
to attempt to access the same data record simultaneously. If they are merely trying
to read the data without updating it, this does not cause a problem. However, if two
or more users are trying to update a particular record simultaneously, say a bank
account balance or the number of available seats on an airline flight, they run the
risk of generating what is known as a '' concurrency problem .'' In this situation,
the updates can interfere with each other in such a way that the resulting data values
will be incorrect. This intolerable possibility must be guarded against and, once
again, the database management system must be designed to protect its databases
from such an eventuality.
A fundamental premise of the database concept is that these three data control
issues—data security, backup and recovery, and concurrency—must be managed
by or coordinated with the database management system. This means that when a
new application program is written for the database environment, the programmers
can concentrate on the details of the application and not have to worry about writing
code to manage these data control issues. It means that there is a good comfort
level that the potential problems caused by these issues are under control since
Search WWH ::




Custom Search