Journal 8


July 2006


Dear Architect,

Welcome to Issue 8 of The Architecture Journal, the theme of which is Data by Design. As architects, I feel we often undervalue the pervasiveness of data in architecture, especially when we consider how data is used in applications and systems that span multiple geographies, time zones, and organizations.

One analogy I frequently use compares the availability of data within a system to water running through the pipes in a house. When we turn on a faucet, we expect clean, filtered water delivered instantly and usually at a consistent, expected pressure. The pipes are the infrastructure in this analogy, and the water is the data. When I think about data and its relationship to architecture, I like to think of the same approach—the data we deliver to users should be clean, filtered, and delivered without delay and as expected—regardless of whether the data is a single e-mail, customer record, or large set of monthly financial data.

Although we will not be covering many plumbing techniques in this issue, we do have a number of great articles from a distinguished group of authors that focus on the importance of data.

We start with Roger Wolter, a solutions architect at Microsoft and author of many white papers and books on SQL Server and SQL Server Service Broker (SSB). Roger discusses the importance of reliability for data, especially in the context of designing connected systems.

Tim Ewald and Kimberly Wolk follow with an article that explains some of the models they created for data integration for the MSDN TechNet Publishing System, the next-generation, XML-based system that forms the foundation of MSDN2. Udi Dahan then takes us on a journey for using entity aggregation to get 360-degree views of our data entities and focusing on concrete ways to solve immediate business needs.

Next, another author team, Tom Fuller and Shawn Morgan, share some of their experiences realizing that data replication can be an antipattern for service-oriented architecture (SOA), especially in light of autonomous services and applications. Then, Dion Hinchcliffe, the CTO of Sphere of Influence, shares some patterns for data consumption and composition, especially in the areas of mash-ups and Web 2.0 applications.

Paul Nielsen ends our series of data-related papers with an overview of Nordic, a new object/relational hybrid model that explores greater flexibility and better performance over traditional relational data models.

Rounding out this issue of the Journal, Andrew Filev shares some of his experience combining an agile methodology with a model of outsourcing, and then Arvindra Sehmi and Beat Schwegler return with Part 2 of their Modeling for Connected Systems series. If you missed Part 1, be sure to download Issue 7 of The Architecture Journal at

Well, that wraps up our theme on data. I trust the articles presented in this issue will help you to design systems with the same data availability as water flowing from the faucet in your house. Of course, I can't guarantee that you won't get your hands wet!

Simon Guest

Articles in This Issue

Reliability in Connected Systems
Loosely coupled, asynchronous, service-oriented applications impose unique reliability requirements. Learn about reliability issues to consider when architecting a connected services application.

A Flexible Model for Data Integration
Organizations use XML data described by XML schema and exchanged through Web services to integrate systems. Find out three causes for failure that data-centric integration projects can encounter, and their solutions.

Autonomous Services and Enterprise Entity Aggregation
Heterogeneous systems manage their own data, which often are not exposed for external consumption. Take a look at how autonomous services transform the way we develop systems to more closely match business processes.

Data Replication as an Enterprise SOA Antipattern
The positives and negatives of data replication can help enterprise architects deliver service-oriented strategies successfully. Discover how to use an antipattern and a pattern to describe data replication for your enterprise.

Patterns for High-Integrity Data Consumption and Composition
The Web is becoming less about visual pages and more about services, pure data, and content. Get acquainted with some patterns that lead to less brittle, more loosely coupled, and high-integrity data consumption and composition.

The Nordic Object/Relational Database Design
An O/R hybrid model provides power, flexibility, performance, and data integrity. Discover how the Nordic Object/Relational Database Design emulates object-oriented features in today's relational database engines.

Adopting and Benefiting from Agile Processes in Offshore Software Development
Offshore outsourcing of software development presents unique challenges. See why modern tools, the global communications infrastructure, and a good offshore partner are critical for agile processes.

Service-Oriented Modeling for Connected Systems: Part 2
Part 1 provided an approach to model connected, service-oriented systems that promotes close alignment between IT solutions and business needs. Now, learn how to implement services mapped to business capabilities.

Download this issue here

This article was published in the Architecture Journal, a print and online publication produced by Microsoft. For more articles from this publication, please visit the Architecture Journal Web site.

© Microsoft Corporation. All rights reserved.