The Internet, data tracking systems, and compliance requirements
is driving data volume growth at alarming rates. On average, data
volumes double every nine months, twice as fast as Moore's Law governing
the growth in computer processing power.
This widening gap between computing power and data growth is compounded
by the increasing need for in-depth analysis, as organisations move
from standard reporting toward interactive ad hoc BI, predictive
analysis and the need for near real-time results.
This rapid growth in database size has to date been managed with
very expensive and consistent upgrading of hardware and software.
However, this trend can no longer keep up with the demands of the
business. In addition, the economic impact of this approach is unfavourable.
Todays businesses need a new approach to data warehousing infrastructure
that is both specific and flexible. Not only must it meet the demands
of the busienss, it must also be easy to deploy and compatible with
the customer's existing BI applications and infrastructure.
Demands of BI Infrastructure
Current State of BI
The current BI infrastructure is a collection of hardware, software
- Database Management System [DBMS] - transaction
processing, holding several hundred megabytes worth of records
with a few internal users. DBMS has been improved in increasingly
complex layers to support terabyte-sized databases and evolving
- Hardware/Operating System - a clustered set
of generic boxes optimized for everything from mathematical queries
to genome investigation.
- Generic File Systems - manage and serve data
for a variety of applications.
Not one of these disparate capabilities has kept pace with database
volumes, complexity or performance demands, in spite of attempts
DBMS grid management - adds another layer to already complex DBMS
Increasingly complex SMP boxes and storage area network [SAN] and
network attached storage [NAS] designed to improve transactional
workloads but have delivered only incremental benefits in the BI
Grid and blade architectural improvements - developed for transaction
systems, and do not meet the demands of BI
Adding more hardware is NOT the answer to growing BI problems.
Doing so requires constant tuning and optimising of user applications.
These systems are continually strained by Terabyte-scale databases.
In addition, such patchwork approaches are extremely difficult
and time-consuming to manage and maintain.
The current solution to this problem is a data
warehouse appliance. In a BI environment, a data warehouse appliance
is a machine capable of retrieving decision-support intelligence
from terabytes of data in seconds or minutes versus hours or days.
They are the only way in which real time or near time decision data
can be made available.
Back To Top
The World's Leading Guide To BI Strategy, Program & Technology
Data Index | Data Defintion
| Meta Data | Data
Management | MDM | Data
Governance | Data Cleansing | Normalization
| Data Integration | Data
Growth | Data Solutions