Database Reference
In-Depth Information
Summary
A distributed database is a single logical database that is physically divided among computers at several
sites on a network. A user at any site can access data at any other site. A DDBMS is a DBMS capable of
supporting and manipulating distributed databases.
Computers in a network communicate through messages. Minimizing the number of messages is important
for rapid access to distributed databases.
A homogenous DDBMS is one that has the same local DBMS at each site, whereas a heterogeneous
DDBMS is one that does not.
Location transparency, replication transparency, and fragmentation transparency are important characteris-
tics of DDBMSs.
DDBMSs permit local control of data, increased database capacity, improved system availability, and
added efficiency.
DDBMSs are more complicated than non-DDBMSs in the areas of updating replicated data, processing
queries, treating concurrent update, providing measures for recovery, managing the data dictionary,
designing databases, and managing security and backup requirements.
The two-phase commit usually uses a coordinator to manage concurrent update.
C. J. Date presented 12 rules that serve as a benchmark against which you can measure DDBMSs. These
rules have local autonomy, no reliance on a central site, continuous operation, location transparency, frag-
mentation transparency, replication transparency, distributed query processing, distributed transaction man-
agement, hardware independence, operating system independence, network independence, and DBMS
independence.
A file server stores the files required by users and sends entire files to the users.
In a two-tier client/server architecture, a DBMS runs on a file server and the server sends only the
requested data to the clients. The server performs database functions, and the clients perform presentation
functions. A fat client can perform the business functions, or the server can perform the business functions
in a thin client arrangement.
In a three-tier client/server architecture, the clients perform the presentation functions, database servers
perform the database functions, and application servers perform business functions. A three-tier architec-
ture is more scalable than a two-tier architecture.
The advantages of client/server systems are lower network traffic; improved processing distribution; thinner
clients; greater processing transparency; increased network, hardware, and software transparency;
improved security; decreased costs; and increased scalability.
Web servers interact with Web clients using HTTP and TCP/IP to display HTML Web pages on Web cli-
ents' screens.
309
Dynamic Web pages, not static Web pages, are used in e-commerce; and server-side and client-side
extensions provide the dynamic capabilities, including the capability to interact with databases.
Cookies and session management techniques are used to counteract the stateless nature of HTTP.
XML was developed in response to the need for data exchange between organizations and due to the
inability of HTML to specify the structure and meaning of its data. XML is a metalanguage designed for the
exchange of data on the Web.
The W3C has developed recommendations for other languages related to XML. These languages include
XHTML, a markup language based on XML and a stricter version of HTML; DTD and XML schema, both
used to specify the structure and meaning of data in an XML document; XSL, a language for creating
stylesheets; XSLT, which transforms an XML document into another document; and XQuery, which is an
XML query language.
OLTP is used with relational database management systems, and OLAP is used with data warehouses.
A data warehouse is a subject-oriented, integrated, time-variant, nonvolatile collection of data in support of
management's decision-making process.
