Presentation on Distributed Database Systems for MITS4003 Course

This presentation provides a comprehensive overview of distributed database systems (DDS). It begins by defining DDS and outlining its goals, which include improving reliability, availability, and performance. The presentation then delves into the different types of DDS, specifically homogeneous and heterogeneous systems, highlighting their characteristics and differences. Key features of DDS are discussed, such as the ability to create, retrieve, modify, and delete distributed databases across various networks, along with data synchronization, transparency, and integrity. The presentation explores factors encouraging the adoption of DDS, such as organizational distribution and the need for real-time data sharing. Advantages of DDS, including modular development, high reliability, and reduced communication costs, are highlighted, along with its disadvantages such as software complexity and the importance of maintaining data integrity. The presentation concludes with references to relevant academic sources.

DISTRIBUTED
DATABASE
SYSTEMS

DEFINITION OF DISTRIBUTED
DATABASES
In a distributed database management
system, the databases do not belong to the
same storage device.
Rather, they are in different devices
connected by a common processing unit.
Distributed databases incorporate a loosely
coupled system.
They are made up of ‘shared nothing
architecture’ [1].
The databases might share the same
network or might belong to entirely different
network.

GOALS OF DISTRIBUTED DATABASE
SYSTEMS
The distributed database systems aims to
improve:
Reliability: Even if one system fails to
function properly in a distributed database,
the other systems can accomplish the tasks.
Availability: The server failure would not
matter, because there will be other systems
to attend client requests.
Performance: Enhancement in
performance is achieved due to the
distribution of the databases at different
locations.

TYPES OF DISTRIBUTED DATABASE
SYSTEMS
Homogeneous DDS
 Two or more database exists in
the network with same DBMS
software
 Data can be accessed and
modified within the same network
 Easy to handle
Heterogeneous DDS
 Two or more databases exists in
the network with different DBMS
software
 Data can be accessed with help of
generic connectivity (ODBC/JDBC)
to modify data
 Complex to handle

FEATURES OF DISTRIBUTED DATABASE
SYSTEMS
• Allows to create, retrieve, modify and
delete distributed databases over
same or different networks
• Responsible for synchronizing
databases to maintain transparency
and integrity of data for the users
• Universal updating of data after
modification
• Maintains privacy and confidentiality
of data

FACTORS ENCOURAGING DDS
 Division of the organizational units
and distribution in different locations
 Need for the support of both OLAP
and OLTP
 To support sharing of real-time data
within no time across units
 To enhance recovery of database
easily
 Readiness to accept different types
of application software within one
network

ADVANTAGES OF DDS
Incorporates modular development across
various sites only by adding computers
and its peripherals, along with a network
Highly reliable because although one
system fails at one site, the other systems
can function until its recovery [2]
Communication cost is much low as
database is shared amongst multiple sites
Response time taken is much less and
better response is achieved due to
availability of the databases at different
locations

DISADVANTAGES OF DDS
Much complex and expensive software
are needed to maintain transparency
amongst data and coordinate across
several sites
Overhead processing is an issue because
simple operations requires
communication to maintain uniformity of
data across sites
Data integrity has to be maintained at a
frequent rate to validate the data
If data is not properly distributed
response time will increase at a large
rate

REFERENCES
[1] G. Alkhatib and R.S. Labban, Transaction Management in Distributed
Database Systems: the case of Oracle’s Two-phase commit. Journal of
Information Systems Education, 13(2), p.4, 2020.
[2] A.A. Sewisy, A.A. Amer and H.I. Abdalla, A novel query-driven clustering-
based technique for vertical fragmentation and allocation in distributed
database systems. International Journal on Semantic Web and Information
Systems (IJSWIS), 13(2), pp.27-54, 2017.