Using Google as a back-end semantic database Oracle Tips by Burleson Consulting May 6, 2007

 

WIP - Not ready for publication

 

The promise of Web 2.0 is being fulfilled and we are now starting to see web "mashups", sophisticated applications that utilize web API's to communicate with disparate information sources.

 

We also see that the web has changed the fundamental nature of access to information.  In traditional database management we see this historical evolution:

 

1. Hierarchical model (IMS) - 1960's

2. Network model (IDMS, IDS, TOTAL) - 1980's

3. Relational Model (Oracle, DB2, SQL Server) - 1990's

4. Object-Oriented model (Oracle Toplink, Ontos, ObjectStore)- 1990's

5. The semantic web (Google mashup) - 2007

Let's take a closer look at the salient features of these software tools as they apply to Web 2.0:

 

IMS Hierarchical DBMS - 1960's - 1970's

 

IMS is a hierarchical linked-list database with one-to-many embedded pointers.  IMS did not naturally represent real-world data relationships (recursive relationships (BOM), many-to-many structures). 

 

Interestingly, even though IMS is a dinosaur database, IMS Fastpath remains the world's fastest engine software, capable of supporting sustained transaction rates in excess of 10,000 per second and running for a decade without interruption.  Check out these statistics on IMS performance and reliability, as of 2002:

 

• Most Corporate Data is Managed by IMS - Over 90% of Fortune 1000 Companies use IMS. 

• IMS Manages over 15 Billion GBs of Production Data

• $2 Trillion/day transferred thru IMS by one customer

• Over 50 Billion Transactions a Day run through IMS.

• IMS Serves Close to 200 Million Users a Day 

• Over 79 Million IMS Trans/Day 

• 30 Million Trans/Day

• 7M per hour handled by one customer

• 6000 Trans/sec across TCP/IP to a single IMS

• 11,700 Trans/sec (Over 1 Billion/day)

• 3000 days without an outage at one large customer
   

Network Database- 1980's

 

The work of the Committee on Development of Symbolic Languages database task group (the CODASYL DBTG) resulted in the Cullinet product the "Integrated Database Management System", or IDMS, and we also see IDS and TOTAL as network model databases.  IDMS was almost completely open-source to their customers, and purchasers of IDMS got DESCT manuals and generous user-exits, making IDMS an extremely powerful engine for large mainframe systems. 

 

IDMS is an extremely flexible two-way linked list structure where you can model complex data relationships such as non-first-normal form records, embedded pointers for object-oriented access. The original IDMS did not support SQL (later, IDMS/R supported SQL access) and relied on "realms" and a navigational data access method that required a programmer to traverse the data structures, unlike SQL which is declarative in nature and build an execution plan on-the-fly.

 

The Relational database - 1990's

 

• Oracle

• DB2

• Informix

• Sybase

• SQL Server

 

Object Oriented Databases - 1990's

 

• Oracle Toplink

• Ontos

• Objectivity

 

Crossover with Dr. Kim's UniSQL object/relational database.

 

The Semantic web

 

• The Google Base project

• The Google API is a programming interface to unstructured (but word-indexed) web pages.

 

 

If you like Oracle tuning, see the book "Oracle Tuning: The Definitive Reference", with 950 pages of tuning tips and scripts.  You can buy it direct from the publisher for 30%-off and get instant access to the code depot of Oracle tuning scripts.