Transcript Informix Red Brick Warehouse

Informix
Red Brick Warehouse
State Of Technology
Rob Totin
Frank Lopinto
Joe Carr
Global Business Environment
Unprecedented dynamics
Constantly changing market conditions
Particularly E-Commerce
Exponential growth in the quantity,
complexity and scope of data
Need for better knowledge management
and analytical capabilities
Get In Front Of The Business
Intelligence Curve
Red Brick Warehouse Empowers Business
with smart solutions for:
data warehouses-open relational databases
web warehouses-web traffic analysis
analytic data marts- “any question, of any
data…Fast!”
INFORMIX
Decision Frontier Solution
Red Brick Warehouse
Proven time-to-market advantages
Specialized server technology
Optimized for:
•
•
•
•
•
analytical data mart solutions
complex queries
fast load performance
high-capacity / high-performance processing
efficient management of very large databases
The Red Brick Advantage
An Independent Analysis
In 1998, Red Brick Warehouse released
version 5.1, claiming that it was the
fastest and most scalable relational
database for data warehousing, data
marts, OLAP and data mining.
The Red Brick Advantage
Skeptics agreed that relational databases
worked fine for data warehouses and data
marts,
However, they questioned the use of
relational databases for OLAP and Data
Mining, believing that:
special functionality was needed for data
mining and
OLAP needed multidimensional modeling to
drill around various dimensions
The Red Brick Advantage
Red Brick Warehouse version 5.1 has
three components:
Database serve - designed to support
databases of 500Gb plus
Load subsystem - transports data, checks
referential integrity and updates indexing in
one integrated run
Gateway technologies - for client / server
access
The Red Brick Advantage
Other design advantages:
Uses compressed indexes to reduce storage
Query partitioning for optimal parallelism
Multiple join algorithms to overcome
performance problems
Uses “hybrid” index types to mix column
domains
RISQL to simplify repetitive DSS queries
The Red Brick Advantage
Red Brick Data Mine Option - for OLAP
and data mining:
The Red Brick premise is: “take the
mining tool to the data instead of taking the
data to the mining tool
Red Brick integrated a neural network,
decision trees and statistical algorithms into
the core of the RDBMS server to allow data
mining directly on the Red Brick
Warehouse database
The Red Brick Advantage
Red Brick Data Mine Option
Users can create multidimensional models
that appear as tables
When data is entered , calculations are
performed and results are stored in model
tables
Tables can be “mined” using RISQL
Tables can be created using GUI-based tools
or extended SQL statements
The Red Brick Advantage
The Red Brick Data Mine approach to
OLAP and Data Mining:
saves considerable time in data extraction,
transformation, shipping and loading data
minimizes redundant storage of data
reduces administrative processing by
consolidating security and admin tasks
mappings between OLAP and warehouse
data are defined in the metedata
The Red Brick Advantage
New Features in Red Brick Warehouse 5.1
Red Brick Vista - enhancements to manage
and process aggregate queries for life cycle
management
Aggregate Advisor - can audit selected
aggregates and compare estimated gains to
actual system usage to calculate cost
formulas for each aggregate. This allows
DBAs to choose which aggregates to create
The Red Brick Advantage
New Features (cont.)
Transparent Query Rewrite - analyzes
complex SQL queries and transforms the
queries to use the appropriate stored
aggregate. Allows administrators to edit
aggregation strategies without affecting
existing applications.
SQL- Backtrack - supports online,
incremental and parallel backups.
The Red Brick Advantage
New Features (cont.)
Table Management utility - a parallel
loader which loads aggregates automatically
when a base table is updated
Red Brick Warehouse Administrator -a
GUI-based tool to control data warehousing
tasks, especially focusing on segmentation
and partitioning.
The Red Brick Advantage
LIMITATIONS
All queries limited to 8k on row size of
intermediate and final result tables
Important because:
• Joining large descriptive columns from the dimension
tables to a wide fact table could exceed this limit
The Red Brick Advantage
LIMITATIONS
The database server has a default stack size
of 5MB
Important because:
• The server will fail if it runs out of stack space
• This may be problematic for data mining operations
The Red Brick Advantage
INDEPENDENT CONCLUSIONS
Red Brick Warehouse’s designs are aimed at
providing:
any data warehouse query
of any complexity
as fast as possible
on very large data warehouses
The Red Brick Advantage
INDEPENDENT CONCLUSIONS
Red Brick Warehouse caters to large data
warehouses:
in data loading and indexing
administrative controls
backup and recovery facilities
all of which promote a high degree of parallelism
The Red Brick Advantage
INDEPENDENT CONCLUSIONS
The Red Brick Warehouse approach to
OLAP is unique
with OLAP functionality built into the relational
database server
most OLAP venders view as a specialized area
needing a multidimensional server
taking OLAP function to the data saves
considerable data duplication and upload/reload
processing
Red Brick Warehouse
Background
1990 - First release of Red Brick
Warehouse
Based on “Star Schema” developed by
Ralph Kimball
Quickly gained support in the field of data
warehousing and data mining
Company fell on hard financial times in
1998
Red Brick Warehouse
The buy-out
Informix, Inc., a world-wide player in the
high-end OLTP market begins negotiations
with Red Brick in the Fall, 1998
By year-end, the deal was done, with
Informix, Inc. paying $35 million for Red
Brick Warehouse
Informix officials were tight-lipped about
intentions for Red Brick
Informix / Red Brick Warehouse
What the deal did for Informix
gave Informix much stronger data
warehousing capabilities
gave Informix new decision-support and data
movement capabilities
gave Informix superior data warehouse
talent
gave Informix significance presence in key
markets
gave Informix the “best in class” in data mart
technology
Informix / Red Brick Warehouse
Between Then and Now
January, 1999 -Informix officials make it clear
that Informix will continue to provide separate
support for its two flagship products, Informix
Dynamic Server and Red Brick Warehouse
July, 1999 - Informix CEO, Dexmier, says
Informix/Red Brick will focus on the Internet as a
revolutionary new market for business intelligence.
Soon to be introduced are new Red Brick
products, i.reach and i.sell, tools to analyze webbased traffic.
Informix / Red Brick Warehouse
Between Then and Now
August, 1999 - introduces Red Brick
Decision Server for advanced analysis of
click-stream data. It supports variable-length
character strings allowing storage of URLs
while minimizing disk space use. Informix is
positioning Red Brick to provide data
warehousing and data mining of web traffic.
Informix / Red Brick Warehouse
Between Then and Now
August, 1999 -Beating all previous results, Red
Brick Warehouse, on a SUN platform, loaded,
queried and scaled a data warehouse to more than
300GB of raw data with up to 600 concurrent users.
Table loading at 14GB/hour was 2.3 times faster than
prior tests
Simulated an environment of 63 stores, 19,000 products,
3.6 million transactions/day and 35 promotions
The data warehouse included two fact tables and five
dimension tables
Informix / Red Brick Warehouse
Where Red Brick Warehouse is Today
Informix will not bury Red Brick Warehouse in
its offerings of Informix products
Informix intends to leverage the name recognition
and reputation of Red Brick Warehouse to the
fullest extent possible
Informix markets Red Brick Warehouse as “an
integral piece of Informix Decision Frontier Solution
Suite”
Informix / Red Brick Warehouse
Conclusion
The global business environment is changing at an
unprecedented rate
The quantity, complexity and scope of data is
growing exponentially
Business must stay in front of the business
intelligence curve
This is why Informix / Red Brick Warehouse
should be the choice for business critical data marts
and data warehousing