|
What does a typical Architect Technology Advisory
Look Like? |
Some Recent ATAs: October 2004: Transactions, Web Service
Style
Several new Web Service standards have been issued related to result
coordination. Mr. Sessions compares these standards. He issues a
prediction that the approach for coordinating results among web services
recommended by both Microsoft and IBM is deeply flawed and that a
different approach will become the de facto standard. The Critical Action
Plan (CAP) addresses how to mitigate the potential for unnecessary code
re-writes. 14 pages
November 2004: Messaging Asynchronous Style
In this discussion of Web service standards related to message delivery,
Mr. Sessions predicts that most organization’s understanding of messaging,
as it will appear in the 2007 timeframe, is incorrect. Because of the
impact this prediction has on overall architecture and the high likelihood
of this prediction, the prediction is rated as 27, or serious. The
resulting Critical Action Plan provides a thorough explanation of what
organizations need to understand about message delivery in 2007 and the
steps they need to take to ensure that their architectures are in synch
with these coming changes. 16 pages
December 2004: Security for Web Services
Most organizations are assuming that the future of Web service security is
based on HTTPS or Kerberos. The December ATA predicts that both of these
assumptions are erroneous, and that systems built on these assumptions
will be obsolete by the year 2007. The resulting CAPs give an extensive
overview of the future direction of security for Web services and how to
plan for the changes that vendors will be putting in place by 2007. 21
pages
January 2005: Enterprise Architectures: A Unified Perspective
Very few organizations have a clear, articulated, universally understood
vision of what a service-oriented architecture looks like, why it is
important, or how applications should plug into it. The January ATA
predicts that more than 70% of the applications written between now and
2006 will need extensive rewrites that could have been avoided with a
clearly articulated vision of Web services, SOAs, and Service-Oriented
Infrastructures. Because of the confidence level of the prediction (.85)
and the code impact (50%), this prediction gets an Alert level of 42.5, or
very serious. The Critical Action Plan describes the web-service, SOA, and
SOI that needs to be communicated across the organization to avoid costly
overwrites. 17 pages
February 2005: Enterprise Architectures: Building Successful Services
Very few of the developers involved in building Web services understand
the special rules that differentiate Web services from other technologies.
Because of this, Mr. Sessions predicts that more than 70% of the
enterprise applications that are built between 2005 and 2006 will require
extensive rewriting to conform to the needs of the SOI, with resulting
costs to a typical large enterprise that will be measured in the millions
of dollars. To mitigate this problem, Mr. Sessions defines the ten
critical rules developers must follow to build successful SOAs. 14 pages
March 2005: Service Oriented Architecture-Fast Track (SOA-FT)
Many organizations are using traditional architectural frameworks, such as
Zachman or TOGAF, to build their SOAs. Mr. Sessions points out that these
frameworks were developed long before the special needs of SOAs were
understood and focused on the needs of building large complex systems
rather than getting enterprise systems to interoperate. Because of the
inherent mismatch of these traditional frameworks to SOA requirements, Mr.
Sessions predicts that most SOAs developed with these frameworks will face
either long delays, cost overruns or both. With a .70 confidence level and
a 70% impact level, this prediction is rated very serious and a CAP
(critical action plan) is issued that includes a streamlined architectural
framework, SOA-FT, that is designed for rapid deployment of high-business
value SOAs. 13 pages
April 2005: Sharing Data in a Service-Oriented
World
Very few organizations understand the special problems of sharing data in
a Service-Oriented Architecture (SOA). Most are assuming that their
traditional approaches to data sharing, i.e. through an shared, enterprise
data base, will continue as they adopt SOAs. There are many compelling
reasons for adopting this approach. However there is one overwhelming
reasons for not adopting this approach: it is almost certain to
fail. Given the costliness of this failure, Mr. Sessions issues a
serious alert level prediction, explaining exactly why the shared
enterprise database approach is fundamentally at odds with SOAs and gives
a detailed Critical Action Plan (CAP) for addressing the proper way to
think about and to share data within an SOA. 14 pages.
May 2005:
Survive and Thrive: The SOA Advantage
Today’s large businesses are typically organized with a brick wall
separating the business folks from the technical folks. Mr. Sessions
believes this is a recipe for failure. Meeting profitability goals and the
new regulatory requirements will demand a close partnership between the
technical and the business sides. SOAs can be the enabling technology for
improving the bottom line while surviving audits, but only if both sides
understand the capabilities and importance of the SOA and Web Service
technology. This ATA starts the dialogue with an explanation of SOAs from
a business perspective. We urge you to share this ATA with the CXOs and
the executive team in your organization to take the first steps toward
creating a strong business/technical
partnership.
June 2005: Visual Studio Team Edition: Who Needs
It?
Microsoft is about to release Visual Studio 2005. This product will
come in a variety of flavors, the most noteworthy of which is Visual
Studio 2005 Team System. Team System is actually a suite of seven related
products. The Team System suite gives an enterprise the ability to
automate the workflow around the software development life cycle. Some
organizations will find Team Suite highly cost effective, as it includes a
host of integrated tools that cover the entire spectrum of software
development life cycle. Other organizations will find Team System suite
overly complex for their needs. Still others will find the suite
hopelessly mismatched to their existing processes. This ATA will help you
decide whether your organization should consider using this new product,
and if so, the steps you should take today to prepare for its introduction
into your organization.
July 2005: Microsoft Solution Framework for Agile
Software Development, Part 1
Microsoft is advocating a process for software development called
Microsoft Solution Framework for Agile Software Development, which I will
abbreviate as MSF. Many people associate MSF with the new Visual Studio
Team System, but MSF can be a great process even for companies that are
not using Visual Studio Team System, or even developing for the Microsoft
platform. MSF is a pragmatic approach for iteratively developing software
systems. It is oriented around the three concepts of business value, team
building and iterative development. This ATA begins the overview of MSF,
an overview that will be concluded in the August ATA. This is your
starting place to see if MSF makes sense for your organization.
August 2005: Microsoft Solution Framework for Agile Software
Development, Part 2
This ATA completes the discussion of the Microsoft Solution Framework for
Agile Software Development 4.0 (MSF). For those of your that decide to
adopt MSF, this ATA will be a valuable jumping off place. But even for
those of you that do not choose to use MSF, a close examination of MSF
will likely reveal gaps in your own process. This ATA introduces the MSF
Fitness Assessment, a self-evaluation approach for judging the fitness of
MSF to your organization. It also looks closely at the work items and work
products identified by MSF, and how the various work streams are broken
down into activities and sub-activities.
September 2005: Orchestrated Web Services
Enterprise Service Busses (ESBs) are becoming increasingly common in large
organizations. They provide a consistent framework for managing workflow
that is highly scalable, robust, and reliable.
Web Services are also becoming increasingly common. They provide a single
approach to interoperability that is platform neutral, industry standard,
and universally supported.
But what happens when these two technologies meet? It turns out that Web
Services as we typically understand them are very poorly suited for
working with a ESB-like architecture. This ATA discusses the merging of
ESBs and Web services, a merging that we call Orchestrated Web Services.
This ATA predicts the problems with which most organizations will meet as
they try to move from a disjoint Web service/ESB strategy to a technology
approach that encompasses both.
October 2005: Why Systems Fail
The Canadian Department of Justice, McDonald’s, and FEMA all have one
thing in common. They have all spent many many millions of dollars are IT
systems that ultimately failed. The root cause of this failure was the
same in each of these cases, and it is the same cause that results in
system failures in hundreds of IT systems worldwide each year at a
cumulative cost of tens of billions of dollars annually.
This ATA starts a two part ATA with a close look at the root cause of some
many failures. Next month we will look at how to prevent these failures.
November 2005: Managing Complexity
The October ATA looked at the complicated issue of complexity. Every
aspect of project management is stressed to the maximum when trying to
deal with large complex systems. The November ATA continues the discussion
of developing a strategy for simplifying complex systems. This strategy
can be applied to any complex project and will reduce risk, accelerate
time-to-value, and increase the chances that complex projects will be
completed successfully.
December 2005: Motion: Microsoft’s New Enterprise
Architecture Framework
Microsoft is moving into a new area: Enterprise Architectures. As part
of this effort, It has been working on a new framework for creating
enterprise architectures. This framework is tentatively named Motion.
Motion addresses several pitfalls in existing architectural frameworks,
but it also has some formidable obstacles to success. This ATA takes a
close look at Motion. It describes what it does well and what it needs to
do to become a success.
January 2006: Why Can’t They Get It? Confessions of a Frustrated
Architect
Companies often spend vast sums of money hiring consulting organizations
to help them make decisions. In almost every case, there are highly
competent in-house staff who could have given the same answers at a
fraction of the cost. Yet these readily available experts are ignored.
The major difference between the expensive consultants and the in-house
experts is communications skills. All too frequently, technical experts
have inadequate training in how to communicate with executives. This ATA
will be a valuable aid to technical experts in how to clearly
communicate ideas to the organization decision makers
.February 2006: The Microsoft Connected Systems
Philosophy
Microsoft has begun a massive reorganization of their enterprise
platform around the vision of connected systems. This reorganization
will impact existing mission-critical products, such as SQLServer and
BizTalk and will be the basis for new products, such as Windows
Communication Foundation and Windows Workflow Foundation. These changes
will have huge impacts on enterprise architectures. This ATA gives an
overview of the philosophy guiding these changes and serves as a
framework for the next four ATAs which will explore how these changes
will specifically impact the areas of security, data storage,
connectivity, and workflow.
March 2006: Communications in a Connected World
This ATA looks more closely at the first pillar of capabilities of
Microsoft’s Connected Systems Philosophy, introduced in the February
ATA. The cornerstone of communications is Windows Communication
Foundation (WCF), the new unified API for all distributed
communications. WCF provides many communications capabilities, but the
most important is Web Services. WCF will be the primary technology used
for creating, designing, and deploying Web Services on the Windows
platform. This ATA goes over the important architectural changes and
shows you what you need to leverage this new technology.
April 2006: Workflow in a Connected World
This ATA looks at the second pillar of capabilities of Microsoft’s
Connected Systems Philosophy: workflow. Microsoft is introducing a new
technology called Windows Workflow Foundation. This will make workflow
easy to use, inexpensive, and readily available. This ATA describes the
new workflow models that are part of Windows Workflow Foundation and
examines the possible impact of this new technology on another product
on which many organizations have bet heavily: BizTalk Server.
May 2006: Security in a Connected World
This ATA looks at the third pillar of capabilities of Microsoft’s
Connected Systems Philosophy: security. While many of the security
features of Microsoft’s systems is well known, Microsoft is introducing
a new user-centric feature code-named InfoCard. This ATA takes a look at
InfoCard, discusses some of the impact it will have on distributed
applications, and discusses some of the hurdles that InfoCard must
overcome to be successful.
June 2006: Data in a
Connected World
This ATA looks at the final pillar of capabilities of Microsoft’s
Connected Systems Philosophy, data. Microsoft has introduced a series of
new capabilities in and around its venerable SQL Server database
intended to better enable services and SOAs. This seems like good news.
And some of those changes are truly going to simplify the building of
SOAs. But some of those changes are going to cause a lot of problems.
This ATA examines each of these changes and recommends which you should
use and which you should avoid like the plague. Don’t plan on using SQL
Server 2005 in an SOA world until you read this ATA and study its
recommendations!
July 2006: Case Study:
Business Process Modeling
In this ATA we begin a new series, the bi-annual case study. We
start by looking at a large retail organization. At first, this
organization’s problems appeared to have technical solutions, but on
closer examination, the problems were almost entirely in their business
process. This ATA discusses an approach to modeling business processes
that we call ABC modeling. ABC stands for autonomous business
capability, and it models the business as a network of connected, but
autonomous business capabilities. This model vastly simplifies the
overall business modeling by ignoring how business capabilities are
implemented, and focusing instead on how they relate to other
capabilities. Once we have an ABC model, we can apply Value Graphs
looking for areas in which we can get the maximum business value for our
investment. This case study shows how ABC models in conjunction with
Value Graphs can help get an overview of an organization, where its
problems lie, and where best to look for solutions.
|
