Database Reference
In-Depth Information
The purpose of a database is to help people keep track of
things. Databases store data in tables in which each table has
data about a different type of thing. Instances of the thing are
stored in the rows of tables, and the characteristics of those
instances are stored in columns. In this text, table names are
written in all capital letters; column names are written in ini-
tial capital letters. Databases store data and the relationships
among the data. Databases store data, but they are structured
so that information can be created from the stored data.
Figure 1-5 lists many important examples of database
applications. Databases can be processed by a single user
or by many users. Those that support many users require
special concurrency-control mechanisms to ensure that one
user's work does not conflict with a second user's work.
Some databases involve just a few users and thousands
of rows of data in a few tables. At the other end of the spec-
trum, some large databases, such as those that support ERP
applications, support thousands of users and include many
millions of rows in several hundred different tables.
Some database applications support e-commerce activi-
ties. Some of the largest databases are those that track users'
responses to Web pages and Web page components. These
databases are used to analyze customers' responses to differ-
ent Web-based marketing programs.
Digital dashboards, data mining applications, and other
reporting applications use database data generated by trans-
action processing systems to help manage the enterprise.
Digital dashboards and reporting systems assess past and
current performance. Data mining applications predict
future performance. The basic components of a database
system are the database, the database management sys-
tem (DBMS), one or more database applications, and users.
Because Structured Query Language (SQL) is an internation-
ally recognized language for processing databases, it can be
considered a fifth component of a database system.
The functions of database applications are to create and
process forms, to process user queries, and to create and pro-
cess reports. Application programs also execute specific ap-
plication logic and control the application. Users provide data
and data changes and read data in forms, queries, and reports.
A DBMS is a large, complicated program used to cre-
ate, process, and administer a database. DBMS products are
almost always licensed from software vendors. Specific func-
tions of a DBMS are summarized in Figure 1-12.
A database is a self-describing collection of integrated
tables. A relational database is a self-describing collection
of related tables. Tables are integrated because they store
data about the relationships among rows of data. Tables are
related by storing linking values of a common column. A da-
tabase is self-describing because it contains a description of
its contents within itself, which is known as metadata. Most
DBMS products carry metadata in the form of tables. As
shown in Figure 1-14, databases also contain indexes, triggers,
stored procedures, security features, and backup and recovery
data.
Microsoft Access is not just a DBMS, but rather an ap-
plication generator plus a DBMS. The application generator
consists of applications components that create and process
forms, reports, and queries. The default Microsoft Access
DBMS product is called the Access Data Engine (ADE),
which is not licensed as a separate product. SQL Server can
be substituted for the ADE to support larger databases.
Enterprise database systems do not combine applica-
tions and the DBMS as Microsoft Access does. Instead, ap-
plications are programs separate from each other and from
the DBMS. Figure 1-16 shows four categories of database
applications: client/server applications, Web applications,
reporting applications, and XML Web services applications.
The five most popular DBMS products, in order of
power, features, and difficulty of use, are Microsoft Access,
MySQL, SQL Server, DB2, and Oracle Database. Microsoft
Access and SQL Server are licensed by Microsoft, DB2 is
licensed by IBM, and Oracle Database and MySQL are li-
censed by Oracle Corporation.
Database design is both difficult and important. Most
of the first half of this text concerns database design. New
databases arise in three ways: from existing data, from
new systems development, and from database redesign.
Normalization is used to guide the design of databases from
existing data. Data models are used to create a blueprint from
system requirements. The blueprint is later transformed into
a database design. Most data models are created using the
entity-relationship model. Database redesign occurs when
an existing database is adapted to support new or changed
requirements or when two or more databases are integrated.
With regards to database processing, you can have one
of two roles: user or database administrator. You may be a
user
of a database/DBMS as a knowledge worker or as an
application programmer. Alternatively, you might be a
da-
tabase administrator
who designs, constructs, and manages
the database itself. The domains of each role are shown in
Figure 1-23, and the priorities as to what you need to know
for each role are shown in Figure 1-24.
The history of database processing is summarized in
Figure 1-25. In the early years, prior to 1970, database pro-
cessing did not exist, and all data were stored in separated
files. The need for integrated processing drove the devel-
opment of early DBMS products. The CODASYL DBTG and
DL/I data models were prevalent. Of the DBMS products
used at that time, only ADABAS and IMS are still in use.
Search WWH ::
Custom Search