‫مالحظة‪ :‬تجدون فصال كامال خاص بهذا العرض‬

‫في مطوية الدروس على منصة موودل‬

‫‪Distributed‬‬
‫‪Databases‬‬
‫‪RABAH MOKHTARI‬‬
Introduction
Distributed database system (DDBS) technology is the union of two
approaches to data processing: database system and computer network
technologies.
1- Database systems
Database systems have taken us from a paradigm of data processing in
which each application defined and maintained its own data to one in which
the data are defined and administered centrally.
2- Computer network technologies
The technology of computer networks, on the other hand, promotes a mode of
work that goes against all centralization efforts.

2
Distributed Data Processing
 Distributed data processing is a computing model in which data
processing is distributed across multiple computers or nodes in a
network.

 The processing can be done in parallel, allowing for faster and more
efficient processing of large amounts of data.

 Each node in the network has access to a subset of the data, and the
nodes work together to process the data and generate the desired
output.

3
Distributed Database system
 A distributed database system is a type of database system that is spread
across multiple computers geographically distributed.

 In a distributed database system, the data is partitioned or replicated

across multiple nodes, and the nodes work together to process queries and
transactions from clients.

 A DDBS is also not a system where, despite the existence of a network, the
database resides at only one node of the network.

4
Distributed Database system

5
DDBS benefits
 Scalability: Distributed database systems can scale horizontally by adding
more nodes to the network. This allows the system to handle large volumes
of data and high transaction rates.

 Fault tolerance: Distributed database systems can continue to operate

even if one or more nodes fail. Data can be replicated across multiple nodes,
so if one node fails, another node can take over without loss of data.

 Improved performance: By distributing the data and processing across

multiple nodes, distributed database systems can improve performance by
processing queries and transactions in parallel.

6
Distributed DBMS architecture
 The architecture of a system defines its structure.

This means that the components of the system are identified, the
function of each component is specified, and the interrelationships
and interactions among these components are defined.

 The specification of the architecture of a system requires

identification of the various modules, with their interfaces and
interrelationships, in terms of the data and control flow through the
system.

7
ANSI/SPARC Architecture
 ANSI/SPARC Architecture is an early milestone in the field of database
systems

 It was developed by the American National Standards Institute (ANSI) and

the Standards Planning and Requirements Committee (SPARC) in the 1970s,
when the field of database management was still in its early stages.

 It helped to establish many of the fundamental concepts and principles that

are still used today.

The ANSI/SPARC architecture defines three levels of abstraction for a

database system

8
ANSI/SPARC Architecture

9
ANSI/SPARC Architecture
 External level: It describes how data is viewed by different users and
groups, and how data is accessed and manipulated by applications. Each
external schema is tailored to meet the specific needs of a particular user or
application.

 Conceptual level: This is the level of the database system that describes
the overall logical structure of the database. The conceptual schema is
independent of any particular application or user, and is used to ensure that
all data in the database is consistent and integrated

 Internal level: This is the level of the database system that describes how
data is physically stored and accessed by the computer system. It defines the
storage structures and access methods used by the DBMS to manage the
data. 10
Architectural Models for
Distributed DBMSs
The ways in which a distributed DBMS can be architected can be classified in
terms of: the autonomy of local systems, their distribution, and their
heterogeneity.

11
Architectural Models for
Distributed DBMSs
Autonomy
Autonomy refers to the distribution of control, not of data. It indicates the
degree to which individual DBMSs can operate independently.
 The local operations of the individual DBMSs are not affected by their
participation in the distributed system.
 The manner in which the individual DBMSs process queries and optimize
them should not be affected by the execution of global queries that access
multiple databases.
 System consistency or operation should not be compromised when
individual DBMSs join or leave the distributed system.

12
Architectural Models for
Distributed DBMSs
Distribution
 Distribution refers to the distribution of data over multiple sites.

 There are two alternatives classes: client/server distribution and peer-to-

peer distribution (or full distribution).

Heterogeneity
 Heterogeneity refers to the presence of diversity or differences in a
distributed database environment in terms of data models, query languages,
and transaction management protocols.

13
Client/Server architecture
 Client/server DBMSs entered the computing scene at the beginning of
1990s and have made a significant impact on both the DBMS technology and
the way we do computing.

 the functions are divided into two classes: server functions and client
functions.

 This provides a two-level architecture which makes it easier to manage the

complexity of modern DBMSs and the complexity of distribution.

 We can cite many examples of DDBMS that use client/server architecture of

distributed database systems. One such example is Microsoft SQL Server,
Oracle Database, MySQL and PostgreSQL.

14
Client/Server architecture

15
Peer-To-Peer architecture
 After a decade of popularity of client/server computing, peer-to-peer have
made a comeback in the last few years as an alternative to distributed
DBMSs.

 Apache Casandra DBMS represent a good example of peer-to-peer

DDBMS and makes use of an entirely peer-to-peer architecture.

 All nodes in a Cassandra cluster can accept reads and writes

16
Distributed query processing
 Distributed query processing is the process of executing a database query
that involves data stored on multiple nodes or servers in a distributed
database system.
When a query is submitted, it must be broken down into smaller subqueries
that can be executed on different nodes in parallel.
 The results must be combined to form the final result set.
 Distributed query processing involves several steps, including query
optimization, query decomposition, data fragmentation and
distribution, data transfer, local processing, and result consolidation.

17
Distributed query processing
The goal of distributed query processing is to minimize the amount of data
that needs to be transferred between nodes and to maximize parallelism in
the execution of subqueries in order to improve query performance .

Query processing problem

 The main function of a relational query processor is to transform a high-
level query (typically, in relational calculus) into an equivalent lower-level
query (typically, in some variation of relational algebra).
 The low-level query actually implements the execution strategy for the
query and The transformation must achieve both correctness and
efficiency.

18
Distributed query processing
Query processing problem
 The main function of a relational query processor is to transform a high-
level query (typically, in relational calculus) into an equivalent lower-level
query (typically, in some variation of relational algebra).
 The low-level query actually implements the execution strategy for the
query and The transformation must achieve both correctness and
efficiency.
Since each equivalent execution strategy can lead to very different
consumptions of computer resources, the main difficulty is to select the
execution strategy that minimizes resource consumption.

19
Distributed query processing
Query processing problem (Example)

20
Distributed query processing
Query processing problem (Example)
following simple user query: “Find the names of employees who are managing a
project”.

The expression of the query in relational calculus using the SQL syntax is

21
Distributed query processing
Query processing problem (Example 1)
Two equivalent relational algebra queries that are correct transformations of the
query above are:

It is intuitively obvious that the second query, which avoids the Cartesian
product of EMP and ASG, consumes much less computing resources than the
first, and thus should be retained.

22
Distributed query processing
Query processing problem
 In a centralized context, query execution strategies can be well expressed in an
extension of relational algebra
 The main role of a centralized query processor is to choose, for a given query,
the best relational algebra query among all equivalent ones.
 In a distributed system, relational algebra is not enough to express execution
strategies. It must be supplemented with operators for exchanging data between
sites
 In addition to the relational algebra operators, the distributed query processor
must also select the best sites to process data, and possibly the way data should
be transformed.

23
Distributed query processing
Query processing problem (Example 2)
 We consider the following query

 We assume that relations EMP and ASG are horizontally fragmented as follows

24
Distributed query processing
Query processing problem (Example 2)
 Fragments ASG1, ASG2, EMP1, and EMP2 are stored at sites 1, 2, 3, and 4,
respectively and the result is expected at site 5.
 Two equivalent distributed execution strategies for the above query are possibles.

25
Distributed database design
In the design of a distributed DBMSs, the distribution of applications involves
two things
 The distribution of the distributed DBMS software, and
 The distribution of the application programs that run on it

Two major strategies that have been identified for designing distributed
databases
The top-down approach and the bottom-up approach

26
 Distributed database design
Top-down approach

27
 Distributed database design
Distribution design

28
 Fragmentation alternatives
Vertical and horizontal fragmentation

29
 Correctness Rules of
Fragmentation
Completeness

30
 Correctness Rules of
Fragmentation
Reconstruction

31
 Correctness Rules of
Fragmentation
Disjointness

32