Distributed Database Management System

1
CHAPTER 10

• Distributed Database Management System

2
1970s

• MIS System
• Support for structured information needs.
• formal reports in standard formats from TPS data.
• 3GL response to precisely channeled request.
• Centrally stored corporate data--mainframe.
• Data access through dumb terminals.
• No ad hoc information for decision makers in a
  dynamic business environment.

3
1980s

• DSS
• Business operations more decentralized
  geographically.
• Global competition.
• Customer demands and market needs--decentralized
  management style.
• Rapid technological change microcomputers.
• LANs became the basis for computerized solutions.
• applications based on DBMSs--data sharing
  attractive.

4
1990s and beyond--DBMS in Dynamic Business
Environment

• Quick ad hoc data access.
• decentralization of management structure and
  business units made decentralized multiple-access
  and multiple-location databases a necessity.
• demand for applications based on data access from
  different sources at multiple locations.

5
What Is A Distributed DBMS?
Figure 10.4 Fully Distributed Database Management
System
6
Distributed DBMS

• A fully distributed database management system
  (DDBMS) governs
• the storage and processing
• of logically related data
• over interconnected computer systems
• in which both data and processing functions
• are distributed among several sites.

7
Distributed Processingand Distributed Database

• Distributed processing shares
• the databases logical processing
• among two or more physically independent sites
• that are connected through a network.

• Distributed database stores
• a logically related database
• over two or more physically independent sites
• connected via a computer network.

8
Functions of a DDBMS

• Application interface
• Validation to analyze data requests
• Transformation to determine requests components
• Query-optimization to find the best access
  strategy
• Mapping to determine the data location

• I/O interface to read or write data
• Formatting to prepare the data for presentation
• Security to provide data privacy
• Backup and recovery
• DB administration
• Concurrency control
• Transaction management

9
DDBMS Advantages

• Faster data processing
• Growth facilitation
• Data are located near the greatest demand
  site.
• Faster data access

• Reduced operating costs
• Improved communication
• User-friendly interface
• Less danger of a single-point failure
• Processor independence

10
DDBMS Disadvantages

• Security
• Complexity of management and control
• Lack of standards

11
DDBMS Protocol

• Transaction processor (TP) receives and processes
  the applications data requests.
• Data processor (DP) stores and retrieves data
  located at the site. Also known as data manager
  (DM).

12
Distributed DB Transparency Features

• allowing the end users to think that they are the
  databases only users.
• manage a physically dispersed database as though
  it were a centralized database.
• Three Levels of Distribution Transparency
• Fragmentation transparencyuser doesnt know db
  is partitioned
• Location transparencyuser specifies location
  name but doesnt know where location is
• Local mapping transparencyend user must know
  both location name and physically where the
  location is

13
Transaction Transparency

• Any transaction will maintain distributed db
  integrity and consistency
• Remote transactionrequest to database can
  reference data at only 1 site
• Distributed transactionrequest can reference
  data from multiple sites but each single request
  gets data from only 1 site
• Distributed requestcan request data from
  multiple sites and each request can be from
  multiple sites

14
Distributed Concurrency Control

• Uses 2-Phase Commit to allow transaction to get
  data from multiple sites
• DO-UNDO-REDO Protocol with Write Ahead
• DOperforms operation records before and
  after in transaction log
• UNDO uses log to reverse transaction if needed
• REDOuses log to REDO transaction if needed
• WRITE Aheadtransaction log written to permanent
  storage before actual data is changed
• Prepare to Commit is sent to all affected
  databases
• Final Commit

15
Data Fragmentation

• Strategy to decide how database will be stored
• Can be at database level or table level
• Table Fragmentation
• Horizontal Fragmentation SELECT subset
• Vertical Fragmentation PROJECT subset
• Mixed Fragmentation combination

16
Data Replication

• Storing multiple copies of data on the network
• Must ensure data consistency
• Affected by database size and usage frequency
• Mutual Consistency Rule
• All copies of data fragments must be identical

17
Replication Scenarios

• Fully replicated
• Multiple copies at multiple sites
• High overhead to maintain
• Partially replicated
• Multiple copies of some fragments at multiple
  sites

18
Design IssuesDistributed Processing

• Must be a division of processing
• Presentation of data
• Manages GUI, user-initiated actions, timers
• Validation
• Consistency of database and user inputs
• Business logic
• Business functions
• Workflow
• Completion of business processes
• Data Access
• SQL statements, queries and modifications

19
Process Management

• Middleware
• Handles client/server communication withouot
  regard to platform
• Oracle--JDBC