My Profile
Active Members
TodayLast 7 Days
more...
Awards & Gifts
Online Exams
Fresher Jobs
Our fresher job section is exclusively for fresh graduates! Find jobs for freshers in major Indian
cities including Bangalore, Chennai, Hyderabad, Pune or Kochi
Resources
Find educational articles, blogs, discussion threads and other resources.
Colleges
Find details about any college in India or search for courses.
|
A national level Technical student’s symposium on TRENDS IN OODB
Posted Date: 14 Feb 2008 Resource Type: Articles/Knowledge Sharing Category: General
|
Posted By: arjun Member Level: Bronze
Rating:  Points: 5
|
|
|
A national level Technical student’s symposium
Technomist-2k8
TECHNICAL PAPER PRESENTATION
ON
TRENDS IN OBJECT ORIENTED
DATA BASE
BY
R.ARJUN
ABSTRACT
The main goal of Object-oriented Database is to capture most of the world’s data that is SQL DBs are designed for simple scalar types, there are more data types text, video, image, audio, geographical data, time series, graphs. In this we will see how the relational model has been extended to support advanced database applications. Features proposed in third-generation database system manifestos. RDBMS currently dominant database technology with estimated sales $50 billion with tools sales included, and growing rate possibly 25% per yr. OODBMS market still small, with sales of $150 million in 1996 and a 3% market share in 1997. OO features being added include user extensible types, encapsulation, inheritance, and polymorphism. Selected features proposed are a 3rd generation DBMS must have a rich type system, inheritance is a good idea, functions, including database procedures, methods… are a good idea.
In this paper we describe MEGA TRENDS in OODB include Commoditization, Packaged Applications, Data Warehouses, Object-Relational Databases, OO DBs, Meta data Repositories.
OOBDMS is different from ORDBMS. In OODBMS the application retrieves relatively few (large) complex objects and works on them for a long period before moving to the next object. But in ORDBMS the application processes a large number of short-lived transactions (e.g. ad-hoc queries) on complex data items.
CONTENTS:
Ø Objectives
Ø Introduction
Ø OO features
Ø The third-generation database manifestos
-The third-generation database system manifesto
-The third manifesto
Ø Mega Trends:
-Commodization of servers
-Packaged Applications
-Data Warehouses
-Object-Relational Databases
-OO DBs
-Metadata Repositories
Ø Conclusion
Ø References
Objectives:
• How the relational model has been extended to support advanced database applications.
• Features proposed in third-generation database system manifestos.
• Object-oriented features in SQL3
Introduction to O-R database systems:
· RDBMS currently dominant database technology with estimated sales $50 billion with tools sales included and growing rate possibly 25% per yr.
· OODBMS market still small, with sales of $150 million in 1996 and a 3% market share in 1997.
· Some expect OODBMS market to grow at over 50% per year, but unlikely to overtake RDBMS
· Vendors of RDBMSs conscious of threat and promise of OODBMS.
· Agree that RDBMSs not currently suited to advanced database apps,
· Reject claim that extended RDBMSs will not provide sufficient functionality/be too slow to cope adequately with new complexity.
OO features being added include:
– user-extensible types
– encapsulation,
– inheritance,
– polymorphism,
– dynamic binding of methods,
– complex objects including non-1NF objects,
– object identity,
– Subtypes and sub tables.
However, no single extended relational model.
The third-generation database manifestos:
Ø The first way: the “3-G Database System Manifesto” (to extend SQL and RDM):
Ø Selected features proposed by CADF:
· A 3rd generation DBMS must have a rich type system.
· Inheritance is a good idea.
· Functions, including database procedures, methods… are a good idea.
· DBMS assigns unique identifiers for records only if no user-defined PK
· Rules (triggers, constraints) will become a major feature in future. They should not be associated with a specific function or collection.
· All programmatic access to a database should be through a non-procedural, high-level access language.
· Should be at least two ways to specify collections, one using enumeration of members and one using query language.
· Updateable views are essential.
· Performance indicators should not appear in data models
· For better or worse, SQL is “intergalactic data speak”.
The second way: the “Third Manifesto”(to preserve RDM and replace SQL) :
Ø proposed by Darwen and Date (1995,2000), it attempts to defend the relational data model:
· Acknowledged that certain OO features desirable, but believe features are orthogonal to RDM.
· Thus, RDM needs ‘no extension, no correction, no subsumption, and, above all, no perversion’.
· However, SQL is unequivocally rejected as a perversion of model.
· Instead a language called D is proposed.
· Primary object is domain - a named set of encapsulated values, of arbitrary complexity, equivalent to data type or object class.
· Domain values referred to as scalars, manipulated only by means of operators defined for domain.
· Both single and multiple inheritances on domains proposed.
· Nested transactions should be supported.
Mega Trends:
• Commodization of servers
• Packaged Applications
• Data Warehouses
• Object-Relational Databases
• OO DBs
• Metadata Repositories
Commoditization:
• Microsoft SQL Server and Access are the drivers
– All low end TPC numbers use MS SQL Server
• Ease of use is critical (e.g. Access)
– automated storage mgmt, DB design, …as easy as files
• Every small business has a DB server
– There will be millions of them
– Or will they rent space on mega-servers?
• Big effects on the high-end DB Business
Packaged Applications:
• Packaged applications drive a lot of DB business
– Little custom application development at F500
• Commercial applications - finance, manufacturing, distribution, human resources, order processing ...
– SAP ($3.8B), People soft ($700M), Baan ($500M), Oracle apps (~ $500M)
– Growing at 40-60% … huge piece of the s/w business
• Mail / bulletin board (Exchange, Domino, …)
aren’t yet on commercial relational DBMSs, but …
Data Warehousing:
• Business goal - decision support for everyone
• Problem - Ad hoc query on production DBs doesn’t work
• Approach - Create snapshot DB for decision support
= a data warehouse
Production Databases
Data Warehouse Data Marts Users
Data Warehouse - Base Technology:
Ø Creating a data warehouse is hard
• Protocol interconnect
• Remote data access
• DB design
• Data scrubbing
• Periodic refresh.
• Query tools
• TPC D is the standard benchmark
Ø a mix of queries (products, suppliers, customers, …)
• Database size
Ø Commonplace - 100 gigabytes
Ø Bleeding edge - 10 terabytes
Data Mining:
• Extract patterns and models from the data
– Automatically, from large data sets
• Techniques
• Statistics
• Visualization
• Pattern recognition
• Machine learning
• Applications - finance, fraud detection, astronomy, market analysis, medical diagnosis, biology …
• A small market, with much projected growth
Object-Relational Databases:
• Goal - Capture more of the world’s data
– SQL DBs are designed for simple scalar types
• More data types - text, video, image, audio, geographical data, time series, graphs, ...
• Richer operators - specialized to data types
– content-based retrieval, image similarity,
spatial contains/overlaps, paths in a graph, ...
• Rules - richer constraints and triggers
– support workflow applications
OR Databases – Examples:
• Query
– Sports cars in accidents within 10 miles of home
– Poorly commented C++ modules implemented last year
– Short docs about data mining reachable from http://foo
• Storage
– albums of digital photographs
– all of a person’s medical records
– all information about an insurance claim
– all information about a product
OR Databases – Technology:
• Adding a new data type affects every DBMS component
• storage structure
• indexing
• algebraic operators
• query optimizer
• query language
• locking
• logging
• streaming/synch
• Components must be designed for extensibility
• Rich type system - nesting, inheritance, relationships
Object-Relational Products:
• All vendors are working on OR DBMSs
– Leaders - Informix, IBM DB2, Oracle 8
– 3rd parties will supply data blades / extenders / cartridges
– Significant features are available today
– But systems are still incomplete and hard to extend
• Plenty of applications to accelerate demand
– Probably the next high-growth market segment
– But not driving lots of revenue yet
Object-Oriented Databases:
• New commercial applications are object-oriented
– Impedance mismatch to store data in relations
– Especially true for design apps: CAD, CASE, CAM
• Stores objects with behavior, not just records
– But OR is competing in this space
J Success factors:
– Efficient client-side objects
– Smooth language binding
– Cache client objects across transactions
L Failure factors:
– Weak TP, SQL query (needs OR), and data dictionary
– Must either replace SQL DB or use SQL DB as its server
– Needs more function to support design apps
D Not a big commercial success so far
– Largest vendor (ODI) is $45M/year growing at 15%
Metadata:
• Scientific data management
• Computer-aided design
-discrete manufacturing
-process industries
• Data warehouse design and management
• Workflow design and management
• Document management
• Heterogeneous query processing
• Application configuration management
• Application development
• More than the past, the commercial market defines the research agenda
– Research in the ‘90s
– Data warehousing and data mining is hot
Comparing OODBMS and ORDBMS:
Advantages:
• Resolves many known weaknesses of RDBMS.
• Reuse and sharing:
– Reuse: from ability to extend server to perform standard functionality centrally.
– Increased productivity for developer and end-user.
• Preserves significant body of knowledge and experience gone into developing relational applications.
Disadvantages:
• Complexity.
• Increased costs.
• Proponents of relational approach believe simplicity and purity of relational model are lost.
• Some believe RDBMS is being extended for what will be a minority of applications.
• OO purists not attracted by extensions either.
• SQL now extremely complex.
OODBMS vs. ORDBMS:
• Use an OODBMS:
– When your application retrieves relatively few (large) complex objects and works on them for a long period before moving to the next object
– When you are happy programming in an OO language
• Use an ORDBMS:
– When your application processes a large number of short-lived transactions (e.g. ad-hoc queries) on complex data items.
Conclusion:
• DB research has been enormously successful
– It drove relational, OO, and OR technology
– Also transactions, distribution, replication
• DB research is a mature field
– Most current work is incremental on long-standing topics
– But there’s much to do, due to changing requirements
References:
[1] http://www.microsoft.com/repository-Philip A. Bernstein
[2] http://www.currenttrendsoodb.com-Chris Clack
[3] http://www.portal.acm.org
 Attachments
TECHNICAL PAPER PRESENTATION ()TECHNICAL PAPER PRESENTATION (9345-14238-ABSTRACT.doc)TECHNICAL PAPER PRESENTATION (9345-14241-ABSTRACT.doc)
|
Responses
|
No responses found. Be the first to respond and make money from revenue sharing program.
|
|
Watch TV Channels
|