SOA Case Study

Case Study

A Sustainable
eBusiness Platform
Using Service Oriented Architecture (SOA)

To comment on this paper, or for more information about CMGC’s service offerings, please contact us at:
info@cmgc.net or on our Web site, www.cmgc.net/feedback.htm

Case Study Version: 1.0
SOA eBusiness Platform Architecture Date: 1/20/2008

Table of Contents
1. EXECUTIVE SUMMARY ................................................................................................................................4

2. INTRODUCTION TO THE VISION ...............................................................................................................6
2.1 KEY CONCEPT...............................................................................................................................................6
2.2 BUSINESS VISION ..........................................................................................................................................8
2.3 TECHNOLOGY VISION ...................................................................................................................................9
2.4 GOVERNANCE VISION .................................................................................................................................10
3. BUSINESS GOALS AND BENEFITS ............................................................................................................10
3.1 GOALS ........................................................................................................................................................10
3.2 BENEFITS ....................................................................................................................................................11
4. ARCHITECTURE STRATEGIES AND REQUIREMENTS ......................................................................12
4.1 ARCHITECTURE STRATEGIES ......................................................................................................................12
4.2 ARCHITECTURE REQUIREMENTS .................................................................................................................12
5. GUIDING PRINCIPLES .................................................................................................................................13

6. CONCEPTUAL ARCHITECTURE VIEW ...................................................................................................14
6.1 DESKTOP .....................................................................................................................................................17
6.2 PORTAL .......................................................................................................................................................18
6.3 PRESENTATION LAYER................................................................................................................................19
6.4 BPM AND BUSINESS PROCESS SERVICES ....................................................................................................19
6.5 APPLICATION-SPECIFIC COMPONENTS/SERVICES .......................................................................................21
6.6 COMMON BUSINESS SERVICES ....................................................................................................................22
6.7 UTILITY/BUSINESS SERVICES......................................................................................................................22
6.8 DATA ACCESS LAYER .................................................................................................................................22
6.9 COMMUNICATIONS LAYER ..........................................................................................................................23
6.10 REGISTRY SERVICES ...................................................................................................................................23
6.11 SERVICE BUS ..............................................................................................................................................25
6.12 SECURITY INFRASTRUCTURE ......................................................................................................................28
7. DEPLOYMENT VIEW....................................................................................................................................30
7.1 DEPLOYMENT ARCHITECTURE VIEW ..........................................................................................................30
7.2 RUN-TIME GOVERNANCE ............................................................................................................................30
8. NON-FUNCTIONAL CONCERNS ................................................................................................................31
8.1 PERFORMANCE............................................................................................................................................31
8.2 AVAILABILITY ............................................................................................................................................32
8.3 RELIABILITY ...............................................................................................................................................32
9. TECHNOLOGY ...............................................................................................................................................32

10. GOVERNANCE ...........................................................................................................................................33

11. SUMMARY...................................................................................................................................................34

APPENDIX A: GLOSSARY.................................................................................................................................35

©Cypress Management Group Corp, 2008 www.cmgc.net Page 2 of 35


Table of Figures
Figure 1. Key SOA Terms...................................................................................................................... 15
Figure 2. Service Operations .................................................................................................................. 15
Figure 3. eBusiness SOA Platform Architecture.................................................................................... 16
Figure 4. Web Services Technology Stack............................................................................................. 17
Figure 5. Portal Architecture .................................................................................................................. 18
Figure 6. BPM Modeling Tool ............................................................................................................... 20
Figure 7. Service Bus ............................................................................................................................. 26
Figure 8. Event Driven Architecture ...................................................................................................... 27
Figure 9. eBusiness Security Scenarios.................................................................................................. 29



1. Executive Summary
Agility, competitive advantage, operational efficiency — all are strategic Company goals. The
objective of the transformation plan outlined in this document is to support those goals and to
enable Company to maximize and sustain its market leadership.
Keys to achieving these goals, and to realizing maximum sustainable benefits from IT operations,
include:
• Alignment with the business
• Agility to evolve quickly to capitalize on new opportunities (translates to time-to-
market)
• Ability to scale appropriately to accommodate growth — in new business volume
and new business features
• Improving ROI on IT investments
• More effective leverage of IT assets
In this document, we define a strategy for transforming the Order Management (OM) platform
and, ultimately, Company’s enterprise IT environment by implementing a services-based
eBusiness platform, supported by strong governance and development processes. Best industry
practices, and the professional experience of members of the team who crafted this vision, make a
strong case for the benefits, business opportunities and ROI that result from adopting a services-
based IT ecosystem − commonly know as a Service Oriented Architecture (SOA).

Gartner definition of a Service: As a gross generalization, [from
the business’ perspective] a service is a repeatable task within a
business process. So, if you can identify your business
processes, and within that, the set of tasks that you perform
within the process, then you can claim that the tasks are services
and the business process is a composition of services.

Adopting this strategy means adopting a new IT perspective on the business — a process-centric
view rather than a traditional application-centric view. Business-to-IT alignment becomes easier
to achieve because business processes are implemented as IT “services” designed to deliver
specific functionality — in fact, the alignment becomes a natural consequence of the platform
architecture. This means that changes are implemented as 'business changes' and 'IT changes' at
the same time, so they are aligned.
Traditional applications are a series of business processes, tightly coupled with each other and
with the data required to support those processes, designed to deliver a well-defined outcome.
The sequences in which the application processes are executed, the policies and rules that define
different paths through the sequence are all “hardwired” into a monolithic application structure.
Any change to processes, rules, or policies impact the entire application.
The eBusiness platform defined here turns that monolithic structure inside out and exposes each
discrete business process. New mechanisms are introduced into the environment that enable the
business to reconfigure existing process flows or implement new process flows (business
models), complete with the rules and polices that govern and constrain those models. Maximum
agility!



However, when thinking about a services-based approach, it’s important to consider the
opportunity it presents as part of a larger business model. SOA is not just software or an IT
discipline. The services model can be applied to the entire business by thinking about what the
company does on a day-to-day basis as being composed of a series of repeatable business tasks or
components.
“The component view of a business is very different from the traditional business model of rigid,
hierarchical organizations and static processes. It is a fundamentally new way to view your entire
business model, and it will help you find new ways to enable flexibility and innovation in all
aspects of your business. This componentized business model is more flexible, responsive and
dynamic and hence is the ideal model for the new hyper-accelerated, networked world.” −
Driving Exponential Change, IBM

Best Practice: Successful SOA initiatives deliver more value to the enterprise if
they include the participation of cross-functional teams including business,
development, operations, and security. These key stakeholders may not be directly
involved on a daily basis but it is critical that they learn from early experience and
understand the challenges and benefits of SOA adoption.

In the sections that follow, we describe the vision that guides this initiative and a roadmap to
achieve the vision. In addition to a technology-agnostic architecture blueprint, we recommend
certain key technologies and “mechanisms” that, in the practical context of Company’s
experience and existing assets, are leading candidates for enabling the vision.



2. Introduction to the Vision
Order Management is positioned to lead Company IT on a mission to transform itself into a more
strategic and valuable asset and partner with the business by providing a flexible eBusiness
Platform that can adapt quickly and cost-effectively to new business requirements. Every mission
needs a clear vision of its goals − a vision that motivates and guides those working to achieve it.
The objective of this document is to define that vision.
An example of high-visibility company that recently undertook a similar transformation is
Federal Express. FedEx is widely recognized as a company that has leveraged information
technology in innovative ways to constantly enhance their original model. Yet even though they
are positioned as a role model for effective use of IT for eBusiness, in 2003 FedEx felt the need to
undertake an IT transformation − a three-year business-process and technology-infrastructure
initiative called 6x6 Transformation which included enhancements to their services-based and
event-driven system architecture.
Why did FedEx feel the need for this initiative? According to Robert Carter, FedEx EVP and
CIO, “Because our customer needs have changed, and we think they'll continue to change
exponentially in the next few years… The vision of the 6x6 Transformation initiative is simple:
FedEx IT will dynamically align resources to the corporation's critical priorities while improving
cycle time and return on investment.”
In 2007, Company IT goals are similar to those that drove the 6x6 Initiative at FedEx:
1. Satisfying changing customer requirements
2. Working as trusted partners with the business and clients
3. Offering greater opportunities for IT employees
4. Improving our ability to deliver results quickly
5. Establishing a consistent environment with enterprise-wide IT infrastructure standards
and common processes
6. Simplifying information access
This document describes the vision for transformation of the Company business-IT ecosystem,
and specifically, for the OM eBusiness Platform. It defines the blueprint for an architecture that
will support current business requirements and ensure that we can accommodate future
requirements quickly and with minimal impact to the application systems and platforms.

2.1 Key Concept
At a high-level, we define a business and IT ecosystem consisting of business architecture, a
process and service model, and a service-based IT platform. Because the theme of this document,
and the foundation for the OM vision, is a service-based architecture (also referred to as Service
Oriented Architecture (SOA)), a working definition of this key concept would be helpful.

Gartner predicts that by 2008, more than 60 percent of enterprises will use
SOA as a "guiding principle" when creating mission-critical applications and
processes.



Even though the concept of a “services-based” IT environment is simple, the industry has
struggled to reach consensus on a definition. One of the reasons is because it means different
things depending on your perspective:
• From the point of view of the business, SOA is a set of services that are configured to
form composite applications with dynamic and flexible process flows. Those
processes and services can be exposed to customers and partners, or to other parts of
the organization
• From the point of view of an enterprise architect SOA is an architectural style that
promotes the concepts of business processes and the orchestration of enterprise-level
business services. It is also a set of architectural principles, patterns and criteria
which address characteristics such as modularity, encapsulation, loose coupling,
separation-of-concerns, reuse and composability.
• From the point of view of a project manager SOA is a development approach
supporting highly productive parallel development.
• From the point of view of a tester and/or quality assurance engineer SOA represents a
way to simplify overall system testing.
• From the point of view of a software developer SOA is a programming model
supported by standards, tools and technologies including, but not limited to Web
Services.
One theme is common to all definitions of Service-Oriented Architecture: it is a way to
decompose enterprise IT systems into smaller, more manageable software artifacts called
services. Through service orchestration, these services are assembled to implement enterprise
business processes. It provides a process-centric view of the enterprise rather than the siloed
application patterns that have proven rigid and costly to maintain and change.
A visual illustration of the synergistic components of the vision, as realized through a service-
based platform, might look like the following diagram of closely fitting pieces of a puzzle.



As the first document in a continuum of architecture artifacts that will fully define the eBusiness
Platform, from enterprise architecture through solution design, this document defines the
reference architectures 1 and will explain the reasoning and motivation for decisions related to the
design patterns and technology stack on which the vision is based. It will describe how the
architecture will support Company standards for scalability, reliability, availability, security,
manageability, and other fundamental architectural principles.

2.2 Business Vision
Company’s vision is to provide a powerful and flexible eBusiness Platform, a platform that
provides secure and appropriate access to eBusiness processes and the data that supports those
processes. To ensure that the business has optimum flexibility the platform will be configurable
to connect processes, people, and information within the organization and across organizational
boundaries.
As Company’s business landscape evolves, the ability to launch a robust eBusiness platform from
which to offer new products and services can provide significant competitive advantages.
Enhanced customer self-service will streamline ordering and fulfillment processes. Software as a
Service holds promise as a radical new business model and product offering.
A service-based architecture enables applications to be designed for flexibility, interoperability,
and change. Case studies show business benefits like faster time to market, reduced application
development cost, and more-competitive electronic service offerings − for example, Software-as-
a-Service.

1
A predefined architectural pattern, or set of patterns, partially or completely instantiated, designed, and proven for use in particular
business and technical contexts, together with supporting artifacts to enable their use.



The extent to which business requirements are fulfilled is directly related to the accuracy with
which business process flows and logic are expressed and automated by the business’ application
software ( see, 6.4 BPM and Business Process Services). Because applications have traditionally
been designed to address immediate or tactical requirements, there is always a challenge in
keeping applications in alignment with business needs when the requirements and direction of the
business changes.
Service-based computing introduces a design paradigm that promotes abstraction on many levels.
One of the most effective means by which functional abstraction is applied is the establishment of
service layers that accurately encapsulate and represent business models. By doing so, common,
pre-existing representations of business logic (business entities, business processes) can exist in
implemented form as physical services.
Because services are intrinsically interoperable, configurable and replaceable, this approach
directly facilitates business change. As business processes are augmented in response to business
climate changes − new competitors, new policies, new priorities − services can be reconfigured
into new compositions that reflect the new business requirements. The architecture evolves in
alignment with the business.

2.3 Technology Vision
The technology vision is a robust and flexible platform that is linked to and aligned with business
strategy and is also modular and capable of adapting to rapidly changing requirements at the
lowest possible cost and highest possible quality. Company’s services-based (SOA) eBusiness
platform is the foundation for realizing this vision. Defining characteristics of the platform
follow. 2
1. The new OM platform is based on a distributed component architecture. Service components
are transparently located inside (or outside) the enterprise and are accessible as services
through a stack of universally supported, interoperable remote procedure call and messaging
protocols. On the wire protocol interoperability − as opposed to code portability − is the
centerpiece of SOA component interactions, enabling universal access and platform
independence. Callers are unaware of a component's underlying implementation technology.
2. SOA components encapsulate functionality and enable reuse at the level of abstraction
modeled by business analysts in formal business models. This improves alignment by
mapping more directly between an IT function and the business function it supports.
3. Declarative, machine-processed contracts enable processes to access the services that an SOA
component provides. These contracts explicitly state functional characteristics as well as non-
functional (quality-of-service, or QoS) capabilities and requirements. SOA components
document their operations using WSDL. Sequences of operations on a component are defined
using Business Process Execution Language (BPEL).
4. SOA components can be dynamically found, selected, and bound by means of their
declarative properties, and integrated using composition mechanisms to form complex
business processes.
Our roadmap will enable us to reach the vision by implementing the new platform in increments –
with significant business value added at each incremental milestone. According to a Forrester
Research finding, “Although you need the guidance of a strategic vision for SOA, it is not wise to
build a strategic SOA platform in a top-down, strategic build-out fashion. Instead, you should

2
Some content provided by: IBM, SOA programming model for implementing Web services



evolve your SOA platform a step at a time based on current day business needs — and do this
simultaneously with ongoing development of your strategic vision.”

2.4 Governance Vision
Sustaining the business and technology visions requires a clearly defined, rational set of
governance processes, with management sponsorship and support to provide accountability and
control. Architectural governance is one aspect of enterprise IT management, and includes:
• Architectural oversight, via a COE or steering committee specifically chartered to
Establish and enforce architectural standards and procedures
Track and monitor new technologies
Provide training and mentoring
• Architecture blueprints and standards
• Technology standards
• Architecture-driven SDLC that includes mentoring and participation by project
architects
A key to establishing enterprise architectures that evolve over time and continue to support the
business’ changing environment and requirements is a feedback mechanism that accepts input
from both the business and IT. Although most of our guiding principles (see 5 Guiding
Principles) are immutable, the actual realization of architectural patterns and components will
certainly evolve over time. The process of architectural governance must be designed to facilitate
that evolution.
Other relevant domains of IT governance include:
• Corporate/Business (to establish goals and strategies)
• Solution Development
• Operations
• Support
Because of the depth and breadth of the topic, specific, detailed governance policies and
procedures are documented separately.

3. Business Goals and Benefits
OM’s architectural approach is driven by a number of overarching business goals. Achieving
these goals will result in significant benefits for the business and for IT development and
operations.

3.1 Goals
Because an application solution is of value only if it supports the business’ goals and mission, the
OM architecture is designed to support the following business goals:
• Increase business and IT agility − Enable Company to meet evolving business
requirements, seize new opportunities, and maintain our position as industry leader
• Business-model focus (process view) − Choreograph and orchestrate processes and
policies that implement innovative and competitive business policies and models



• Reduce incremental spending
• Achieve increased ROI
• Drive IT investments based on business objectives
• Simpler and more effective IT governance
• Improve process efficiency − Enable more efficient process flow with a clear view of
all orders, from all sales channels, through the sales and fulfillment lifecycle
• Improve quality of application releases
• Leverage existing assets and investments − Leverage high-value software and
infrastructure assets as components for the new OM platform.
• Reduce operational overhead

3.2 Benefits
In addition to the benefits that the business will realize as a result of achieving the goals outlined
above, the architectural approach defined in this document will deliver the following benefits:
• Improved time-to-market for new business models − As described in Section 6, the
new platform will enable definition and deployment of new and revised business
models much more quickly than the current OM environment
• Reduced application lifecycle cost (Total Cost of Ownership) − A component and
service-based architecture will result in a significant savings in costs for new
development and on-going maintenance
• Higher productivity − Development responsibility can be assigned based on the
functionality that each component provides, and the skill sets required to implement
that functionality. Implementation of new features is faster because changes are
isolated to the component to be replaced
• Higher quality − Defects, defined as missed requirements and errors in application
logic, are dramatically reduced as a direct result of the architectural approach and
supporting SDLC processes
Requirements are easier to understand and implement
Encapsulation of logic into components ensures that the impact of change to a
component is isolated to that component. Other aspects of the system are not
exposed to the risk of new defects
Developers are able to focus on specific components of logic responsible for the
logic or process being change d or implemented
Full test coverage is easier to attain
• Longer useful life − Because the system can more easily evolve over time as new
functionality is required
• Reuse of common functionality − Locally, and on an enterprise scale
• High availability and reliability − Enabled by deployment of redundant components



Company will realize these benefits – and others -- because a component-based architecture is
based on fundamental design principles, referred to as “separation of concerns” and “loose
coupling.” Separation of concerns dictates that each component is designed to provide a specific
type of service (a specific responsibility). Loose coupling refers to a design approach that
minimizes the dependencies between components.

4. Architecture Strategies and Requirements
The OM architecture will enable Company to implement application functionality and platform
infrastructures that provide a robust, reliable, and high-performing solution.

4.1 Architecture Strategies
The strategies that will enable the OM project to achieve its goals and realize its benefits include.
• Component-based. The architecture is defined as a collection of components that
are responsible for specific, distinct business functions.
• Service Oriented. Service-Oriented Architecture is an architectural style that
physically separates functionality into services that can be used by multiple
application and system functions.
• Plugable architecture. By using service-based architecture and open standards for
integration mechanisms and protocols, business functionality can be added, changed
or removed without impacting the system overall. Thus, the design allows the
business community to alter the basic nature of the functions performed without
changing the overall architecture and infrastructure of the systems.
• Data abstraction. By providing mechanisms to define metadata, the solution space is
isolated from changes occurring in the virtual data-models of the order and
fulfillment systems. The same meta-data and universal data-dictionary are used in
the access, manipulation and presentation of data and hence will provide a much
higher level of flexibility to the business community as a whole.

4.2 Architecture Requirements
The architecture of the OM platform provides the context within which the application is
developed, and defines the guidelines and constraints within which we will make design and
implementation choices. The architecture, therefore, establishes the technical requirements for
the final realization of the platform. Requirements for the OM platform include:
• Scalability. Establishes the ability of an application or platform to provide additional
capacity over time. Adding capacity will often require the addition of resources, but
should not require changes in the architecture, design, or implementation. There is
no inherent limit to the scalability of an application implemented in compliance with
this architecture. The platform can be scaled horizontally (e.g., to accommodate
increases in user load) and vertically (e.g., to accommodate greater processing
complexity).
• Performance. The OM solution will meet its required Service Level Agreements
(SLA). The OM architecture ensures that application code will be implemented in a
way that will enable the solution to meet SLA performance requirements. The
application architecture addresses performance concerns by enabling the most
effective use of caching, connection pooling, object granularity and coupling, and
other techniques that improve throughput and reduce latency.



• Reliability. This attribute is a measure of the reliability of the individual underlying
components. System reliability describes the likelihood of component failures at any
level − application, server, or network. The architecture supports platform
configurations that provide redundant, highly available processes.
• Availability. The modular and redundant characteristics of the OM architecture will
enable the system to maintain a high percentage of availability. Most system
maintenance can be performed without shutting down the application.
• Security. The scope of security requirements extends from the edge of the solution
domain through fine-grained application functions. It includes the levels of
authentication supported, the granularity of role-based authorization, the mechanisms
for auditing, the techniques used to ensure data integrity, and the resistance of
unauthorized access (intrusion detection). Other considerations may include security
credential cache and session timeout, LDAP (Lightweight Directory Access protocol)
performance and administration.
• Flexibility. The flexibility of a system is the extent to which the system can be
modified to meet the changing needs of a business. To be flexible a system has to be
simple (in design, not necessarily in functionality), consistent, and modular
(component-based). Individual components should be capable of being reorganized
to provide new services. Encapsulated service components and careful adherence to
“separation of concerns” at all levels of the solution will help enable flexibility.
• Maintainability. Ensures that the system can be repaired and components replaced,
with minimal impact on availability and level of service. The modular and redundant
features of the OM architecture will enable routine maintenance to be performed in
compliance with the availability requirements, stated above.
• Interoperability. Is dependent on the framework that enables components to work
with each other, and the ability to add new components. The OM architecture is
based on an open, interface- and service-based design. The result is that OM
functionality that may be used by external applications can be made available as a
service to those applications.
• Reusability. Software applications represent a valuable corporate asset. Component
architectures provide an opportunity to create a portfolio of software assets that may
be used by multiple projects or applications. The result is a decrease in application
lifecycle costs and an improvement in productivity and quality.

5. Guiding Principles 3
Architecture principles are used to guide architecture decisions, including project scope, solution
design, software and technology purchases, and deployment options. For example, a principle
stated as, “should conform to enterprise data standards” guides database design. “Should have the
fewest number of physical servers possible” guides deployment decisions. “Should deliver
business value in quarterly increments” guides project scope decisions.
The OM conceptual design complies with the following principles:
• Separation of concerns − Each component is designed to provide a specific service
(a specific responsibility). This principle applies at all levels of granularity. At the

3
For more information about OM architecture guidelines refer to the document, “OM Architecture Principles and Guidelines”



finest level, a component may be specifically concerned with computing some result
from a pair if input values. At a larger granularity of service, the ESB (for example)
is concerned with guaranteed delivery of messages and tracking related events.
Large-grained components or services are usually composites of smaller services.
• Loose coupling − A design approach that minimizes the dependencies between
components. By implementing the ability for a component to locate other
components, whose services it requires, and to discover at run time what interface
protocol is required to use those components, our architecture enforces loose
coupling.
• Process View − Business is viewed as a series of processes that can be
choreographed to form complex business models.
As the OM architecture evolves to detailed levels of design and implementation, we will establish
additional design and implementation guidelines and standards for the project.

6. Conceptual Architecture View
The practice of deriving an information system architecture from a business architecture, using a
process-centric perspective, results in a Service-Oriented Architecture. The business processes
and the tasks from which they are composed can be thought of collectively as services. Thus, the
business design is a composition of services, and the policies and rules that define and constrain
their use, which form the information system design.

Note: Perhaps the most common misconception about Service-Oriented
Architecture (SOA) in the marketplace is that SOA and Web Services are
the same thing. This point of confusion is widespread and affects architects
and developers, consultants and vendors alike.
Our architectural vision includes the use of Web Services where
appropriate but our definition and implementation of “services” is not
dependent on the use of Web Services technology and protocols.

Our view of the OM eBusiness platform is a modular, component and service-based architecture
(SOA) that decomposes the system into discrete components that provide specific functionality.
Figure 2 illustrates some of the terminology use to describe SOA architecture.



Figure 1. Key SOA Terms 4

A guiding principle is “separation of concerns.” Each block in Figure 4 encapsulates specific
concerns (aspects) of the overall architecture.
Another key concept is that of “consumers” and “providers” of services. Service consumers
invoke the services of a service provider, based on a published “contract” that defines how the
service is called and what results (service) it provides. Figure 3 is a classic illustration of the
process of registering, discovering and using a service.

Figure 2. Service Operations

This approach enables systems to be assembled so that when changes or improvements are
required − for example, a better presentation technology is available, or a new business process is
developed − the appropriate service component can be changed with minimal impact on the rest
of the system. The Service Registry clearly defines each service’s interface so that other
components can invoke services without regard to how the component is implemented and how it
functions, internally.

4
Illustration provided by IBM.



The following diagram illustrates the major architectural aspects and suggests their physical
relationship with each other. It provides a high-level architectural pattern. This high-level
pattern is a composite of more detailed architectural patterns that define each individual aspect.
It is important to view this diagram as a representation of architectural mechanisms and
capabilities, not as a physical view of the platform. For example, although there are many mature
products available to instantiate each aspect, Company may elect to use any number of
approaches − open source, commercial or internally developed − to implement any specific
aspect. The significant takeaway from this diagram and discussion is that each aspect is important
to realizing the benefits of the component and services-based approach.

Figure 3. eBusiness SOA Platform Architecture

Company’s reference architecture is based on state-of-the-art standards and technologies that are
considered best practice in implementing service-based systems. The service technologies are
vendor neutral and interoperable. They include a web services framework and associated
technology stack (see following diagram) including:
• Extensible Markup Language (XML) − XML is the markup language that underlies
most of the specifications used for Web services. XML is a generic language that is
used to describe the content in a structured way, separated from its presentation to a
specific device.
• Simple Object Access Protocol (SOAP) − SOAP is a specification for the exchange
of structured XML-based messages between the service provider, service consumer,
and service registry. SOAP, in principle, is a protocol-independent transport.



• Web Services Description Language (WSDL) − WSDL is an XML-based interface
and implementation description language. It describes a service’s “contract.”
• Universal Description, Discovery, and Integration (UDDI) − UDDI is both a client-
side API and a SOAP-based server implementation that can be used to store and
retrieve information on service providers and Web services.
This reference architecture also includes a service bus (ESB) to ensure loose coupling and a
Business Process Management (BPM) engine to provide flexibility and rigor in choreographing
and orchestrating business processes.

Figure 4. Web Services Technology Stack

The following sections describe each aspect of the Business Platform architecture, the benefits
each offers and alternatives for implementing each aspect, along with the tradeoffs associated
with each alternative.

6.1 Desktop
OM’s Desktop environment provides consistent, integrated user access to all OM functions, for
authorized users. This user-facing tier is a Web-based interface that provides a single point of
access to a wide variety of data, knowledge, and services—at anytime and from anywhere using
any Web-enabled client device.

6.1.1 Benefits
Because the Desktop environment is necessary for the delivery of services to users, its benefits
are obvious. There are, however, specific benefits related to each implementation alternative (see
below),

6.1.2 Products and Technologies
Alternative technologies employed to provide the Desktop environment may include thin-client
HTML browser pages or rich client applications, based on an Ajax framework (JavaScript and
XML), which uses local resources and XML Web Services and can be deployed and updated
from a centralized server.



A purely browser-based presentation allows full control of page composition to remain on the
Web server. Load on users’ clients is low because they are only responsible for rendering the
HTML sent by the server. Most user interactions, however, require that the server compose a
new page and send it back to the client. For heavily interactive applications, the result is a
significant amount of network traffic and load on the server.
Rich internet applications (RIA, and an aspect of Web 2.0) like those provided by Ajax, execute
logic on the client − in the form of JavaScript − that can perform many routine tasks locally,
without requiring a round-trip interaction with the server. When a specific piece of data is
required form the server, the messages between client and server are much smaller than a full
Web page, and therefore much faster. Because logic executes locally, applications tend to be
faster and appear more responsive. SLAs for interaction may be easier to achieve and sustain.

6.2 Portal
A web portal aggregates applications and content to allow Company users, partners, employees
and customers to tailor their user experience, with personalized applications based on role,
context, actions, location, preferences and collaboration needs. As the front end to a service-based
architecture, portal software provides a component-based model that supports reusable software
assets. Figure 5 illustrates the high-level architectural view of a typical portal.

Figure 5. Portal Architecture



6.2.1 Benefits
A Web portal interface provides each user with a personalized, unified view of the enterprise and
access to specific business functions − or topics − from a single access point. From an
architectural perspective, it provides a single point of focus for implementing and controlling
consistent and secure access to enterprise assets.

There are numerous commercial portal products. Because Company has invested in XXX’s suite
of products, [XXX], is a logical choice.

6.3 Presentation Layer
Presentation layer services support the interaction between applications and end-users. If
required, presentation logic can orchestrate the interface to non-human interfaces.
Regardless of to what or whom the service is interfacing, every external interaction shares one
thing in common – the ability to project a view of the system tailored to the specific interaction
fidelity, frequency of interaction, and presentation composition that best fits the needs of the end
user or device.
Interactions are tailored to role-sensitive contexts – adjusting what is seen and the behavior
presented to the external world based on whom the user is and what role they are performing.
Authentication, authorization, and proximity may all be significant to what a user can do and
how.

6.3.1 Benefits
Isolating all responsibility for presenting information to the user allows the business to change to
the way in which the information is presented − flow, format, validation, etc. − without requiring
changes to the underlying business logic. Conversely, it also allows changes to business logic and
data models without impacting the users’ view.

OM currently uses a Struts MVC framework, one of the original web-application frameworks and
the most pervasive JEE MVC framework deployed today. An alternative gaining wide acceptance
is the Spring Framework. Both Struts and Spring are open source products.
The two viable options for use in the new architecture are:
• Struts 2.0
• Spring MVC
Struts and Spring can be used as complementary frameworks − can co-exist. If OM decides to
implement a Spring framework, we will define a roadmap that enables both frameworks to exist
until the re-implementation is complete.

6.4 BPM and Business Process Services
This aspect (or, component) of the architecture provides the realization of “applications” from the
vertical view − a set of processes that enable the business to perform a related flow of processes.
This new process-centric paradigm maximizes the business’ ability to define new, and redefine
existing, business models to accommodate virtually any orchestration of processes necessary to



support evolving strategies. In the event that a new model requires new processes, those processes
are implemented as services and plugged into (wired into) the process flow.
The following diagram illustrates an example of a typical BPM process modeling UI.

Figure 6. BPM Modeling Tool

Process services include various forms of compositional logic – the most notable of which are
business process flows and business state machines. Both kinds of composition mechanisms, plus
other forms such as business rules and decision tree processing may be to be equally valid
approaches to choreographing service composition.
Business processes are governed by rules which determine how tasks are routed, which task will
be executed next, and how long the process can wait for a response from another system or
human being. BPM requires two types of management:
• Operational process management
• BAM (business activity monitoring)
The former provides a mechanism for dealing with exceptions, understanding the current state of
processes and dealing with in-flight as well as long-running asynchronous processes; the latter
provides the means by which both business and process efficiency is managed, optimized and
understood.

6.4.1 Benefits
A BPM suite enables business users and business analysts to collaborate to model process flows
and the rules and policies that govern that flow. Analysts and developers build forms and



processes, simulate process execution, compare simulation results with production data,
“playback” the model for testing, and optimize the process.
In a study conducted by the Aberdeen Group in 2006, “Business Process Monitoring systems are
achieving an average 9% improvement in revenues. Other key business value metrics
enterprises have achieved by using Business Process Monitoring systems include a median 12%
decrease in process-related expenses, a 10% return on assets (ROA) and an 18% return on
investment (ROI). Some 94% of respondents’ companies are either moderately or very pleased
with the overall improvement Business Process Monitoring has delivered.”
Some specific benefits relevant to Company include:
• Optimization. A significant BPM benefit is the optimization of business processes
across departments and even cross-enterprise. There is a direct correlation between
more efficient business processes and specific cost savings.
• Business Agility. Another area where BPM provides a significant return is in
improved agility. Business parameters and market conditions change rapidly, and
managers need to be able to respond to those changes quickly and effectively. Using
BPM, will allow Company to create “on-demand” business environments we can
adapt quickly and easily at the business level to build a competitive edge.
• IT Agility. Experience with SOA is demonstrating that, as an enabler to respond to
business climate changes, by adapting technology infrastructures to be more
responsive, SOA is a best-practice approach to building and sharing applications that
run the business. With SOA, Company can build components, or services, of
applications that business models will reuse and share across the enterprise to bring
cohesiveness to business process.
At Company, specific business requirements and strategic objectives dictate the need for some
form of business process management. An evolving, complex multi-channel business model, with
complex relationships and variable policies, is one example.

Because of assets already in place, key candidates for the choice of a BPM product for OM
would be:
• XXX
• YYYYYYY
Other potential candidates include . . .

6.5 Application-Specific Components/Services
This aspect of the architecture contains functionality specific to a Company business model
(function), implemented as services. An example of a business-specific function implemented in
this aspect is GOLD. Application components are not, generally reusable. They contain logic
specific to one business function or business model.
These components are created specifically as services within a business model and represent the
basic building blocks of business logic – services that are not decomposable within the business
model, but that can be composed to form higher-level services.



6.5.1 Benefits
These services may be thought of as the “root” of an application model − defined at the Business
Process Services layer (6.4 BPM and Business Process Services). Providing these components as
services enables flexible definition of new business models, and enables the business to define
(via the BPM) composite applications that consist of multiple, logically related business models.

As with all of the service layers, these services may be implemented using Web Service
technologies, Java RMI protocols or other appropriate mechanisms.

6.6 Common Business Services
The common business services aspect contains business-centric services and bases its functional
boundary and context on one or more related business entities. These are highly reusable services
because they are agnostic to most parent business processes. As a result, a single entity service is
leveraged to extend the functionality of multiple application-specific business processes.
Examples of services in this category might include licensing, pricing and access to CPM.

6.6.1 Benefits
Encapsulation and isolation of logic; loose coupling; re-use.

technologies, asynchronous messaging, or Java RMI protocols.

6.7 Utility/Business Services
It is beneficial to establish a set of utility services that are not business-centric. This results in a
distinct, technology-oriented service layer. Utility services are dedicated to providing reusable,
crosscutting utility functionality, such as event logging, notification, and exception handling.
They are generally application agnostic in that they can consist of a series of capabilities that
draw from multiple enterprise systems and resources, while making this functionality available
within a very specific processing context.

6.7.1 Benefits
Encapsulation and isolation of logic; loose coupling; re-use.

technologies, asynchronous messaging, or Java RMI protocols. In specific cases, a utility
software product may be deployed as a service at this layer.

6.8 Data Access Layer
The Data Access Layer provides an abstraction between the physical data sources required by
OM, and the application code that uses that data. The physical structure of the data source is
separated from the business logic so that the impact of changes to either the physical database or
to the business processing logic is minimized.
In this SOA architecture, the data access layer is defined as a data “service.” Doing so allows OM
to make its data available to any consumer − internal service or external application − as



appropriate. Definition of a canonical data model is a best practice to support the abstraction of
data services.

6.8.1 Benefits
A primary benefit is isolation of the physical data structure from consumers of the data. Services
or applications are provided a “view” of data the correlates with their intended use.
Providing access to data via encapsulated services also enables effective and flexible security
policies. Access and authorization is controlled at the source.
Data is, of course, a critical corporate asset. Providing “data as a service” maximizes the use and
value of that data and enables the business to leverage the capability to enhance its strategic
value.

One responsibility of the data access layer is object-relational mapping (ORM). OM is currently
considering two of the leading products in this domain:
• Hibernate − an Open Source project and a component of the JBoss Enterprise
Middleware System (JEMS) suite of products. Hibernate is one of the most widely
used ORM frameworks.
• TopLink − a commercial product from Oracle. TopLink was one of the first ORM
products on the market.

6.9 Communications Layer
The OM solution is required to send and receive data to a variety of external “end points.” To
allow the application to focus only on the data that is to be transmitted, and the logical
destination, the Communications Layer takes responsibility for transforming communication
requests into physical communications protocols, and manages the communications sessions until
the transmission is complete.
This architectural approach enables either end of the physical communication to make changes to
the means of communication without impact on the application.

6.9.1 Benefits
Primarily a separation-of-concerns benefit − responsibility for preparing messages for physical
transport and routing, as well as selecting protocols compatible with the endpoint are isolated in
this architectural component.

The choice of specific communications products will be finalized during the detailed design
phase and will probably be selected from the technology stack provided by the platform − e.g.,
WebSphere.

6.10 Registry Services
Governance and the need for a point of control within the SOA environment demands that service
information artifacts be stored and managed in a consistent manner that allows enforcement of
organizational policies. This is the role served by a registry service in an SOA deployment.
The following example describes a registry-repository using a library metaphor:



• A registry-repository is like a local library.
• It has a repository that contains all types of electronic assets, much like the library
bookshelves.
• It has a registry that contains metadata describing the electronic artifacts, like the
library’s card catalog contains information describing the published content on its
bookshelves.
• The registry and repository are administered jointly. In a library, the card catalog
information and books on the shelves are administered jointly.
• Across the enterprise, any number of registry-repositories should be able to work
together to offer a federated, unified service, much like multiple libraries can
participate in a cooperative network and offer a unified service.
An SOA registry service should provide governance capabilities that enable organizations to
define and enforce organizational policies governing the content and usage of the artifacts
throughout their life cycles. Since organizational policies vary − e.g., security and authorization −
an SOA registry-repository should enable organizations to enforce custom policies for the
governance of any type of service throughout its life cycle. In particular, it should enforce
conformance to such policies when a service information artifact is published or updated.
Most service registries comply with UDDI (Universal Description Discovery and Integration
protocol) protocols. The following description is taken from the Oasis UDDI Specification.
“For Web services to be meaningful there is a need to provide information about them
beyond the technical specifications of the service itself. Central to UDDI’s purpose is the
representation of data and metadata about Web services. A UDDI registry, either for use
in the public domain or behind the firewall, offers a standard mechanism to classify,
catalog and manage Web services, so that they can be discovered and consumed.
Whether for the purpose of electronic commerce or alternate purposes, businesses and
providers can use UDDI to represent information about Web services in a standard way
such that queries can then be issued to a UDDI Registry – at design-time or run-time –
that address the following scenarios:
• Find Web services implementations that are based on a common abstract interface
definition.
• Find Web services providers that are classified according to a known classification
scheme or identifier system.
• Determine the security and transport protocols supported by a given Web service.
• Issue a search for services based on a general keyword.
• Cache the technical information about a Web service and then update that
information at run-time.
These scenarios and many more are enabled by the combination of the UDDI
information model and the UDDI API set. Because the information model is extremely
normalized, it can accommodate many different types of models, scenarios and
technologies. The specification has been written to be flexible so that it can absorb a
diverse set of services and not be tied to any one particular technology. While a UDDI
Node exposes its information as an XML Web service, it does not restrict the technologies



of the services about which it stores information or the ways in which that information is
decorated with metadata.”

6.10.1 Benefits
As an indication of the importance of a service registry in service-based platforms, a Gartner
opinion in 2006 stated, "In 2006, lack of working governance mechanisms [e.g., registry] in
midsize-to-large (greater than 50 services) post-pilot SOA projects will be the most common
reason for project failure (0.8 probability).“
Primary benefits include:
• Enables and supports reuse of services across the enterprise. By enabling discovery
and self-describing contracts for usage, a service registry makes the goal of software
asset reuse a reality
• Provides a mechanism for governance of asset usage and security.
The registry is a central element of the service-oriented approach to software design. By enabling
policy-based distribution and management of enterprise Web services, a registry delivers
significant business value.

Generally, service registries are embedded in other mechanisms, like application servers, ESB
and workflow engines. Some specific registry products include:
• IBM WebSphere Service Registry and Repository (WSRR)
• Systinet Registry − A mature and widely used SOA registry
• SAP Services Registry
• Sun Service Registry
The choice of a service registry for Company will probably depend on other components of the
technology stack.

6.11 Service Bus
The enterprise service bus (ESB) is a silent partner in the SOA logical architecture. Its presence
in the architecture is transparent to the services of your SOA application. However, the presence
of an ESB is fundamental to simplifying the task of invoking services – making use of services
wherever they are needed, independent of the details of locating those services and transporting
service requests across the network to invoke services wherever they are deployed in the
enterprise.
An ESB provides an infrastructure that removes the direct connection dependency between
service consumers and providers. Consumers connect to the bus and not the provider that
implements the service. This connection further decouples the consumer from the provider. A bus
also adds value by providing capabilities such as security and guaranteed delivery. It is an
advantage to implement these capabilities centrally within the bus at an infrastructure level rather
than within the application. The primary driver for an ESB, however, is that it increases
decoupling between service consumers and providers.
In the following diagram, an ESB is depicted as a logical component in a service-based
architecture as the mediator between the service consumers and service providers. The service



providers and service consumers never interact directly. The ESB provides services to resolve
differences in protocol and format, and decouples the service consumer from the service provider.

Figure 7. Service Bus

A service bus supports implementation of an “event driven architecture” (EDA). EDA defines a
methodology for implementing applications and systems in which events transmit between
software components and services. EDA complements SOA by enabling long-running
asynchronous processes. An EDA node (application or service) posts events to the ESB and does
not depend on the availability of consuming service service. It is decoupled from the other nodes.
EDA is often referred to as "event-driven SOA".



Figure 8. Event Driven Architecture

An ESB is an architectural pattern and can be implemented by many different products across the
enterprise, and assembled together to act as a federated enterprise bus. Many vendors are now
offering a complete product to fulfill enterprise integration needs and there are several open-
source enterprise-capable products.

6.11.1 Benefits
A service bus enables faster, simpler, and more flexible integration, which enables the business to
implement services quickly. It reduces the number of interfaces and improves the reusability of
interface components to cut cycle time from design to deployment.
An ESB provides features that improve responsiveness, customer service, transaction time, and
partner interactions. An ESB provides capabilities that enhance both direct connection between
applications and routing requests between applications.
An ESB supports the concepts of a services-based implementation by:
• Decoupling the consumer’s view of a service from the implementation of a service
• Decoupling technical aspects of service interactions
• Integrating and managing services in the enterprise
Decoupling the user’s view of a service from the implementation increases the flexibility of the
architecture. It allows substitution of one service component for another (for example, if a new
service is required to fulfill new business requirements) without the user being aware of the
change or without the need to alter the architecture to support the substitution.
From a practical, business perspective, what are the benefits of an event-driven architecture?
Users view virtually all activities as “events” − contract in place, order completed, fulfillment
complete are a few obvious examples. Enabling the system to respond automatically, based on
policy and workflow rules, provides a powerful process governance and support capability.



At the system level, event generation and monitoring enable run-time monitoring and control to
support consistent service levels. For example, we may define a five-second threshold for
response to a user request. In the event that those transactions begin to fail to complete within the
specified SLA, an event can be generated to cause a reallocation of resources.

Service buses are one of the most mature and long-standing of the architectural mechanisms
define in this architecture. Some obvious candidates for Compnay’s use include:
• XXX
• MULE − Open source product from MuleSource
• WSO2 − Open source product based on Apache Synapse and Apache Axis2 projects
• YYYYYY Fabric ESB
It’s worth noting that there appears to be a new generation of light-weight, high-performance
service bus products emerging.

6.12 Security Infrastructure
Security concerns in the first generation of OM have primarily focused on user authentication
(GSSO) and authorizations (Entitlement). As OM evolves to an eBusiness platform, with
external customer-facing features and business-to-business capabilities, additional dimensions of
security begin to surface.
In a traditional security model, a system's security mechanisms, such as a firewall or virtual
private network (VPN), screen out unauthorized (human) users. An externally facing e-business
platform, however, requires that the architecture be open to access from multiple systems and
external non-Company users. If services are exposed without an appropriate security mechanism,
a hacker can configure a machine to impersonate an external partner or customer’s system and
make fraudulent service calls.
The following diagram illustrates the two aspects of eBusiness security − human to machine
interaction and machine-to-machine interaction.



Figure 9. eBusiness Security Scenarios

Security is an issue in service-based environments because the SOA stresses machine-to-machine
interaction, while most IT security is based on human-to-machine interaction. Authentication and
authorization become more challenging in this environment. It is virtually impossible to block
unauthorized use of web services in an unsecured SOA. It is relatively easy to intercept a message
in an unsecured SOA and reroute it or transform its content for illicit purposes.
There are solutions to SOA security that can address all of these issues. As we determine an
appropriate level of security − during the physical architecture design − we will consider
implementing a security solution that enables SOAP message monitoring; federated
authentication; application proxy; contract management; certificates, keys, and encryption; and
audit logging.

6.12.1 Benefits
The benefits of a secure eBusiness environment are best described in terms of risk management
and containment. No commercial enterprise would want to launch a Web-based platform, and
particularly a customer facing platform with rigorous security in place.

Implementing appropriate levels of security for OM may include:
• Public/private key encryption (PPK)
• Encrypted XML messages. SOAP messages used to communicate with web services
are encrypted to ensure privacy.
• Security Assertion Markup Language (SAML) − an assertion (token) that expresses
the authenticity of the user in a way that will be accepted by the web service that the
user is invoking.



7. Deployment View
A key benefit of a component-based system is that it enables great flexibility in deploying
software components and services to achieve maximum performance and resilience.

7.1 Deployment Architecture View
We will provide a physical deployment view as we define the physical architecture, in subsequent
releases of this document. The architectural approach defined herein enables a great amount of
flexibility in how all aspects of the architecture are implemented and how they may be deployed..

7.2 Run-time Governance
Componentized and services-based environments are most effective when monitored and
controlled by an effective run-time governance infrastructure. There are numerous commercial
products available that provide extensive features for ensuring that the run-time environment
stays healthy and performant.
Governance can allow the business to establish flexible SLA policies to cover a variety of
circumstances. For example, on an eBusiness platform it may be desirable to ensure that external
users (customers) are given priority in the use of resources. The governance infrastructure can
react to degrading performance by throttling internal users’ access and giving priority to
customers.
Some other driving requirements for a run-time monitoring and control include:
• Visibility − What services are deployed in the environment
What services, SOA components, and infrastructure are installed and in use
Who is using them
Are they meeting expectations / SLAs
• Control – Taking action to prevent or correct Issues…
Define and enforce runtime policies – make sure proper policies are active
Diagnose failures / prevent them
• Ensuring Integrity – Ensuring changes do not impact the application environment…
Automatically check for the correctness of the running system
Detect and validate changes before they impact users and partners
• Exception Management − Identify and react to exceptions in real-time…
Transaction tracking and failure events
Alerts when processes or transactions do not complete
• Enforce Service Level Agreements − Dynamically take corrective action…
Enforce agreements based on business criteria, e.g., “Gold” users, Accounting
systems at the end of quarter
Activate more resources (blade servers, load balancing, etc.)
Redirect traffic
Make lower-priority users wait until high-value users complete transactions



Successful, scalable management of SOA requires the ability to take action and intervene to
ensure the run-time environment is under control.

8. Non-Functional Concerns
In addition to the functional aspects addressed by this architecture, there are non-functional
requirements that any system or platform must meet. These requirements must not be lost or
subsumed to the functional elegance of the architecture. They are equally − sometimes even more
− critical to meeting business goals and requirements.

8.1 Performance
Performance is a well-known issue when architecting, and implementing, a service-based
architecture. Early adopters, over-enthusiastic about the obvious benefits that SOA represented,
failed to consider performance early enough (at the conceptual architecture level) and ended up
with SOA implementations that failed miserably because they could not perform, or were
inordinately costly because the hardware infrastructure had to be scaled considerably beyond the
budgeted levels. As often happens in the early stages of the adoption of new technologies, those
failures created a general impression that it is too expensive and complex to implement a high-
performance SOA architecture.

8.1.1 Latency
Transaction latency is a primary focus area when designing and implementing services-based
solutions. The nature of componentized software applications introduces more potential
bottlenecks in the end-to-end flow of data from source, through transformations and processing,
and to the consumer (user). Depending on the specific source of latency, there are a number of
techniques to reduce or virtually eliminate delays.
When the performance cost of loosely coupled components is too high, tighter coupling may be
appropriate. Although counter to the principle of loose coupling, there are situations when a
compromise can significantly improve performance. One example is avoiding the use of the
service registry to “find” a service. If the service definition can be “known” in advance, overhead
of invoking the service may be much lower (also see, Service Granularity).

8.1.2 XML Overhead
XML has emerged as the standard language of service-oriented infrastructures, forming the basis
for all current SOA standards. The performance issues in SOA deployments due to verbosity and
size of XML, combined with chatty transport protocols is an area that requires specific focus.
Different approaches have been adopted to deal with this performance bottleneck, key among
which are:
• XML compression into binary format
• Caching
• Offloading via accelerators (appliances)
• Transport-level optimization
• Optimized XML processing
While the first three approaches listed above are common in commercial deployments in different
products and custom implementations, the latter two − transport level optimization and optimized
XML processing − are still new to the SOA product community.



8.1.3 Service Granularity
Probably the most obvious performance factor in a service-based architecture is related to the
granularity of services. While finer granularity optimizes our ability to orchestrate composite
services, and to change them easily, careful analysis of our service model and its map to the
business process model, often reveals that there is not practical need to decompose certain
functionality to the degree that is physically possible.
As a general guideline, services implemented as Application-Specific services (see, Architecture
Figure 3) are very course grained. These services encapsulate the largest amount of related
functionality − they tend to be the equivalent of “application” systems in an application-centric
view. Services in the next category, Common Business services, are finer grained components
and services. The scope of their responsibility (concern) is much narrower. Utility services are
implemented at the finest granularity; they have a very specific function, usually very narrow in
scope.
The level of coupling often tracks inversely to the granularity. Large-grained business services
are loosely coupled and may be implemented as Web services. The fine-grained components are
accessed as local services, or even from a library.

8.2 Availability
Service-based platforms greatly enhance IT’s ability to provide 24/7 availability. Because
application components and the infrastructure that supports them are location independent and
can be distributed, it’s practical to implemented redundant services, on redundant clustered and
load balanced servers.
When application components need to be replaced, the SOA architecture (specifically, the service
registry) allows component versioning. Consuming applications don’t need to be stopped while
services are updated.

8.3 Reliability
Reliability is not a new non-functional requirement. It has been one of the primary “…ilities” for
years. Service-based architectures introduce some opportunities for enhancing reliability, and
some challenges that must be addressed. Some of the specific characteristics and mechanisms
employed to implement a service-based system that facilitate high reliability, include:
• Service Bus. My decoupling sender and receiver of messages, and providing
guaranteed delivery (via persistent queues) the ESB prevents loss of transactions and
protects against the failure of any specific node.
• Redundant services, on clustered servers ensure that the physical failure of any single
node, or network segment, does not cause the system to fail.
A cautionary note is that, this architectural approach only enables deployment of a highly
available platform. IT must provide the appropriate hardware and network infrastructure to
realize the availability benefits.

9. Technology
At the conceptual level, the OM architecture is technology and product agnostic. In practice,
however, there are obvious best practices and cost-effective approaches for implementing each
architectural feature. For example, we could choose to implement our own service bus for
messaging and event management − using technologies like Java, database triggers, and JMS.
This “home grown” approach would not only be a more costly and risky implementation choice,



it would result in a long-term maintenance issue. The better choice is to choose one of the proven,
supported products − commercial or open source − available off the shelf.
The following table defines key technologies that comprise the SOA stack, and lists potential
candidates for each.

Item Description Tradeoffs

Details deleted to honor confidentiality agreements

10. Governance
Service Oriented Architecture requires a shift in the way software is developed and deployed. The
paradigm shifts from “Develop now, Integrate later” to a “Develop for Integration”. This new
paradigm, combined with the technologies and standards created to support this shift, requires
SOA to be implemented in a well-planned, well coordinated, and effectively managed way.
To ensure business continuity, reduce integration costs and complexities, limit corporate
liabilities including security, and effectively compete in the marketplace, Company will govern
the design, development, deployment, and operations of new services in the enterprise.
Governance is the ability to ensure that all of the independent efforts (whether in the design,
development, deployment or operations of a Service) come together to meet SOA requirements.
In the early stages of introducing an SOA ecosystem, the most effective Architectural
Governance structure is a centralized authority able to provide singular focus on defining and
establishing standards and procedures.
In order to effectively implement and manage the SOA environment, the answers to the following
questions need to be answered and monitored:
• Policies − what are the policies that we have? Where are they implemented?
• Business Models − what are the business models to be developed?
• Business Services − what services, at what level, do we need?
• Compliance Status − to what degree do our services conform to our policies? Which
interfaces are not in compliance? What is the impact of the noncompliance on the
operations of the Services or our business operations (e.g. security breaches, service
levels, etc.)?
• Impact Analysis - what happens to our SOA operations if we change our current
policies? How do we add new policies?
• Interdependencies - how will operations be impacted by changes made to Services?
Which critical processes will be effected or cease to operate?
• Exception Management - can we grant an exception to a defined policy for a certain



project? What will be the impact of that exception?
Governance applies to all stages in the services lifecycle.
• Service design and development practices
• Configuration management
• Release management
• Contract management
• Service monitoring and control
• SLA management
• Runtime mediation
• Incident management
• Change management
There are three levels of design-time governance that should be taken into account:
• Portfolio governance: what initiatives, projects, and deliverables (both services and
applications consuming those services) get funded and the tracking of those projects’
progress against funding and timeline objectives.
• Architectural governance: ensuring that the services and applications developed
under SOA initiatives conform to the organization’s business and technical
architectures and best practices.
• SDLC governance: the day-in day-out application of SDLC best practices (e.g., unit
test before code promotion, peer review of code changes, establish an SCM system
with code promotion levels) when developing services.
Portfolio and SDLC governance have been around for quite some time, and a number of tools
exist to support these governance activities. SOA, because of its loosely coupled nature places
increased importance on architecture governance. If effective architecture governance is not
instituted, there will be a high risk that service spaghetti will result.

11. Summary
The intent of this document is to communicate the architectural vision and to gain consensus with
Company business and technical communities. The document fulfills that intent by
• Documenting the strategic business goals
• Defining the architectural goals
• Providing a conceptual design to support the attainment of those goals
The Architecture Vision document will continue to evolve to present a more detailed view of the
OM architecture and design.



Appendix A: Glossary

Term Definition
Enterprise Service An enterprise service bus (ESB) refers to a software architecture mechanism,
Bus (ESB) implemented by technologies found in middleware infrastructure products based
on standards, that provides foundational services for more complex architectures
via an event-driven and standards-based messaging engine (the bus). An ESB
combines event-driven and service oriented approaches to simplify integration. It
acts as an intermediary layer to enable communication between different
application processes.

Service Registry A catalog or index that acts as the “system of record” for the services within an
SOA.

SAML Security Assertion Markup Language (SAML) is an XML-based standard that
provides a framework for describing security information about a user.

BPM Business Process Management.
SDO Service Data Objects (SDO) provide a framework for the design and use of
business objects in an SOA. The fundamental concept in the SDO architecture is
the data object. SDO is often used interchangeably with the term data object.
SCA Service Component Architecture (SCA) is a model for application development
that splits the application function from the implementation details. SCA defines
modules and components that are connected using standard interfaces.

WebSphere Process WebSphere Process Server is an SCA-compliant runtime element that provides a
Server fully converged, standards-based process engine that is underpinned by
WebSphere Application Server. It is, along with WebSphere Enterprise Service
Bus, a strategic product for integration and modernization of IT assets, including
core systems using SOA


SOA Case Study

Recommended

More Related Content

What's hot (20)

Viewers also liked (10)

Similar to SOA Case Study (20)

SOA Case Study