Three Cheers for Three Tiers
by Angie Arnold and Paul Miller
Today's communications business environment is one of the most
rapidly and dynamically evolving industries in the world. For that
reason, speed and flexibility are important characteristics of successful
service providers. In fact, a key differentiator among providers
is the ability to rapidly deploy and configure their operational
support systems (OSS) to continuously meet the demands of customers,
technologies and the marketplace. The trick, of course, is to do
so cost-effectively.
The underlying architecture of a service provider's OSS is a determining
factor in how well its organization can react to change. A two-tier
architecture, for instance, has inherent scalability and interoperability
problems. It cannot effectively distribute the processing power
between the client and the application server, resulting in a "fat"
client or server. While this is acceptable in a relatively homogeneous
environment with fairly static business rules, it becomes a limitation
when that environment changes and grows and resources cannot be
flexibly allocated. Two-tier structures also entail heavy network
traffic since there is no middle tier to handle multiple server
requests and complex data access. As the business environment grows,
so too do development efforts, migration costs and maintenance workloads.
The architecture of a service provider's
OSS
is a determining factor in
how well it can react to change.
A three-tier architecture, on the other hand, is a flexible and
scalable way of organizing distributed client-server systems. It
clearly separates client programs from data sources, allowing the
programs to be maintained more easily. This enables providers to
cost-effectively match their functional and processing resources
to the needs of their entire organization. The three-tier architecture
model is comprised of three major application layers that support
the following key functions:
- The user layer provides
user interface control, application navigation, application style
and presentation, and interface points for data entry and validation.
- The business layer maintains
business rules associated with application processing logic along
with tables and parameters necessary to control and configure
the OSS based on business requirements.
- The data layer handles access
to data while safeguarding information based on a set of definable
security rules related to users, organizations and external applications
that interface to the core OSS.
This three-tier architecture delivers four key benefits to service
providers: enhanced operational efficiency, ensured data integrity,
increased scalability and reduced deployment costs.
Enhanced operational efficiency
A three-tier architecture enhances operational efficiency in several
significant ways. First, it allows localization of servers to points
of need; service providers can devote a single server to a specific
area of service, thereby guaranteeing adequate capacity for transaction-heavy
processes. Second, it enables providers to distribute discrete portions
of the application across multiple servers based on their organization's
individual business needs, processing requirements and geographical
hierarchy. Finally, three-tier architecture minimizes software distribution
if only server-side portions of the application are touched during
development; this feature reduces or eliminates costly upgrades
related to client-side portions of applications resting on hundreds,
if not thousands, of remote machines.
Ensured data integrity
A three-tier architecture eliminates the time-consuming system
errors caused by fraudulent data and reduces the risk of data corruption
and/or loss that can result in uncontrolled system outages or failures.
It does this by handling changes to data at centralized control
points, ensuring that all user and system requests are properly
formatted when submitted to the database. Changes are written once
and placed on the middle tier to be available throughout the system.
Database restructuring, upgrades, migration and other changes can
be performed without stopping or altering the client programs. Furthermore,
concurrency controls preserve data integrity when multiple application
clients simultaneously perform requests against one or more servers.
Increased scalability
By separating code into "client" and "server" code, a three-tier
architecture increases the scalability of service providers' applications.
They can deploy an application's components and data on one or more
servers based on current organizational and geographical processing
needs. Then they can reconfigure and re-deploy their system resources
as the application's user base grows and data and transaction volumes
increase. In short, service providers that choose an OSS with a
three-tier architecture can be confident that it will readily scale
with their business.
Reduced deployment costs
A three-tier architecture separates business logic from data layers
within the application, thereby giving mission-critical front- and
back-office systems (such as billing and service activation) direct
access to published interface definitions. The ability to expose
both business and data rules to external applications greatly reduces
the time and cost required for initial implementation and for subsequent
upgrades to the core OSS and/or to various external systems connected
to it.
A three-tier architecture eliminates
time-consuming system errors caused by
fraudulent data and reduces the risk
of data corruption and loss.
Once the server-side applications have been completed, the development
time for client-side applications is greatly reduced. This architecture
eases future modification to interface points. They can be isolated
when change is required so as not to affect other internal or external
application functions, thus reducing the risk of system failure
or a complete outage.
The bottom line
The benefits of a three-tier architecture combine to decrease service
providers' costs and increase their revenues. It decreases costs
by maximizing data integrity, reducing wide-scale deployment of
client-side application components, and allowing increased access
at points. This architecture increases providers' revenues by enabling
more users to access a wider variety and larger numbers of local
and remote servers, facilitating rapid implementation with other
OSSs, and eliminating labor-intensive overhead associated with administration
of multiple servers.
Angie Arnold,
PhD, is a writer specializing in telecom at Eftia.
Paul Miller
works in the Office of the CTO at Eftia.
From Current OSS, Winter 2001, Vol. 2, No. 2. Published by Eftia OSS Solutions.
|