David Chappell Blog

Last week, at the International SOA & Cloud Symposium I had the distinct pleasure of participating in a 3 day workshop to produce a SOA Manifesto. Participants were asked in advance to provide their input as prep work, and each of us presented our wisdom and guidance on what we thought should be in such a document. A team of 15 contributors including technologists, analysts, and practitioners came together and debated over a value system and a set of guiding principles. After 3 days of what was sometimes heated discussion, we published the SOA Manifesto.

If you like what you see, you are encouraged to become a signatory. Also, twitter search string is #soamanifesto

The text is included below, for your convenience, along with some of my commenary inline. Also, a video to the unveiling ceremony can be found here.

Preamble: Service orientation is a paradigm that frames what you do. Service-oriented architecture (SOA) is a type of architecture that results from applying service orientation. We have been applying service orientation to help organizations consistently deliver sustainable business value, with increased agility and cost effectiveness, in line with changing business needs.

Through our work we have come to prioritize:

NOTE: The following represents a set of core values we came up with to help guide you through the decision making process during your SOA journey. Both sides of the value statement are things that we think are important. This is not intended to be a set of mutually exclusive things, but simply a value system should a tradeoff over priorities need to be made.

Core Values:

Business value over technical strategy

All too often architects get too caught up in building the perfect framework for their SOA implementations. This is indeed an important goal, but don't let it cause you to lose sight of the business value you are trying to provide. Make sure you are continually revisiting the business benefits of what you are trying to achieve, and make sure you are still in alignment with them.

Strategic goals over project-specific benefits

Subtly integrated into this simple statement is the idea that you need to scope your SOA projects to provide you with a series of success points that can be achieved in a reasonable amount of time, yet accompany your projects with a longer term strategy of where you want to get to. But don't let project-specific quick wins get you into short-cuts that keep you from you longer term strategic goals.

Intrinsic interoperability over custom integration

This means services should be designed to integrate and interoperate as a core competency rather than an afterthought. Even if you are service enabling a legacy system this can be accomplished if it is given a priority upfront as part of the design. NOTE: Using an ESB or SOA infrastructure to achieve this is certainly allowable, whether its part of your design strategy up front or an afterthought.

Shared services over specific-purpose implementations

Most of the time services should be built with sharing and reuse in mind. This may not always be the case, but an organization should strive to create services that are shared and reused across departments and across projects.

Flexibility over optimization

The primary benefit of using service oriented principals is to build flexibility and adaptability to change. This sometimes comes at a cost, so be aware of what the cost of flexibility is. BTW this does not imply that a flexible solution based on SOA can't scale or perform well.

Evolutionary refinement over pursuit of initial perfection

The key word here is initial perfection. SOA is a journey. Service components and their means of interaction are very likely to change over time as the needs of the business change and as you identify holes in your initial implementations. Pursuit of initial perfection is great if you can get it the first time, but don't let it stand in the way of your strategic goals or your project goals.

That is, while we value the items on the right, we value the items on the left more.

I will be providing more insight into the following Guiding Principles over subsequent blog posts -

Guiding Principles

We follow these principles:
Respect the social and power structure of the organization.
Recognize that SOA ultimately demands change on many levels.
The scope of SOA adoption can vary. Keep efforts manageable and within meaningful boundaries.
Products and standards alone will neither give you SOA nor apply the service orientation paradigm for you.
SOA can be realized through a variety of technologies and standards.
Establish a uniform set of enterprise standards and policies based on industry,
de facto, and community standards.
Pursue uniformity on the outside while allowing diversity on the inside.
Identify services through collaboration with business and technology stakeholders.
Maximize service usage by considering the current and future scope of utilization.
Verify that services satisfy business requirements and goals.
Evolve services and their organization in response to real use.
Separate the different aspects of a system that change at different rates.
Reduce implicit dependencies and publish all external dependencies to increase
robustness and reduce the impact of change.
At every level of abstraction, organize each service around a cohesive and manageable
unit of functionality

Here is the full list of Authors:

Ali Arsanjani Grady Booch Toufic Boubez Paul C. Brown David Chappell John deVadoss Thomas Erl Nicolai Josuttis Dirk Krafzig Mark Little Brian Loesgen Anne Thomas Manes Joe McKendrick Steve Ross-Talbot Stefan Tilkov Clemens Utschig-Utschig Herbjörn Wilhelmsen

© 2009, the above authors
this declaration may be freely copied in any form,
but only in its entirety through this notice.

BTW that copyright notice is really intended for anyone to freely requote, or redistribute the content as long as it is in its entirety.

Here is a partial list of blog links from other authors on the team (in no particular order) -
Anne Thomas Manes
Stefan Tilkov
Dave Chappell (moi)
Steve Ross-Talbot
Brian Loesgen
Paul Brown
Mark Little
John deVadoss
Clemens Utschig-Utschig
Joe McKendrick
Ali Arsanjani

Cheers,
Dave

I'm looking forward to Oracle OpenWorld next week

Its always a great time, very informative, and a great way for customers, partners, and Oracle staff to mingle and talk about technology. I would like to invite you all to come to my sessions on Tuesday.

This one will have Carl Dumont, CTO of Wellpoint, co-presenting with me talking about their use of Coherence in their SOA environment to cache results from expensive backend system calls -

ID#: S307456
Title: SOA and Application Grid for Improving Back-end Response Times and Reducing Customer Abrasion
Track: Service-Oriented Architecture and Business Process Management
Date: 13-OCT-09
Time: 11:30 - 12:30
Venue: Marriott Hotel
Room: Salon 12/13

Abstract: Service Orientation allows IT to modularize business applications into reusable service components that can be combined together in new ways that are more flexible and capable of readily adapting to business growth and change. However through service enabling of legacy application assets into reusable business services, you may be unwittingly driving increased traffic toward backend systems that weren't ever designed or appropriately sized to handle such loads. In this session, co-presented by Wellpoint CTO Carl Dumont, we will learn how Wellpoint is using a service result cache to build a Data Grid into their SOA fabric that offloads and dramatically reduces load on backend systems, resulting in a cost savings in millions per year.

ID#: S308618
Title: The Service-Result-Cache Pattern: Using a Data Grid in the SOA Tier
Track: Oracle Develop: Service-Oriented Architecture
Date: 13-OCT-09
Time: 14:30 - 15:30
Venue: Hilton Hotel
Room: Yosemite A

Abstract: By service enabling of legacy application assets into reusable business services, you may be unwittingly driving increased traffic toward backend systems that weren't ever designed or appropriately sized to handle such loads. The challenge of building excess capacity into backend systems to handle peak loads at certain times of the day, month, or year can be costly and time consuming.

An Application Grid makes it possible for a business to meet SLA's by enabling applications to achieve predictable latency under increased sustained loads. When combined with a SOA using the Service Result Cache pattern, an Application Grid can dramatically reduce load on backend systems by caching results from frequently accessed services.

I encourage you to also check these summaries of all the SOA Sessions, SOA Governance sessions, and Application Grid sessions.

I hope to see you there!
Dave

Hi all,
I recently took a cue from Mike Piech, who has been doing some video white board sessions, and decided to do some of my own that discuss how customers are using SOA and Application Grid together to increase customer satisfaction, improve response times for customer facing applications, reduce load on backend systems, and generally build apps that scale with predictable latency. A general introduction on the subject is found here ,
And a discussion of how customers are using the “service result cache” pattern reduce load on backend systems can be found here –

Enjoy,
Dave

Customer Case Study to be featured at Burton Group Catalyst Conference next week.

I have been working with Carl Dumont, the CTO of Wellpoint, for some time now on integrating Application Grid into Wellpoint’s SOA projects. Next week, at Burton Group’s Catalyst Conference in San Diego, I will co-present with Carl on two exciting projects that both involve using Oracle Application Grid in a SOA environment to cache results from service calls to dramatically reduce load on backend systems.

We will present a Cost Benefit Analysis and PoC findings that show 1) Improving response times, reducing down time, and reducing customer abrasion for customer facing applications, and 2) Saving $MMillions in MIPS charge by offloading mainframe usage by integrating caching of results from mainframe service calls in the SOA layer.

Here is the title and abstract for the talk -

Cost-Benefit Analysis: Using SOA and Application Grid to both enhance performance and availability of customer facing applications and reduce mainframe usage charges

The cost of hosting applications on mainframes can run in millions per year. How much of that is allocated on excess capacity and for peak loads and otherwise goes unused? Even for non-mainframe applications, how do you measure the cost of system downtime to your business? What is your customer abrasion factor for self-service apps that are slow to respond and sometimes fail? How can SOA help? SOA, when combined with new advancements in distributed caching and application grid technology can dramatically improve response times for sluggish applications, and also dramatically reduce the cost of mainframe MIPS. How do you explain this to the business?

This session, presented by a CTO who wrestles with these issues every day, will discuss a Cost/Benefit Analysis and Proof of Concept that takes us through 2 use cases: 1) Measuring the "customer abrasion factor" resulting from unresponsive self-service applications, and using SOA and Application Grid technology to dramatically improve customer satisfaction. 2) Quantifying the cost of excess capacity for mainframe hosted applications, and Using SOA and Application Grid to dramatically reduce mainframe MIPS usage.

The conference runs Wed - Fri July 29 - 31, and our session will be Wed at 3:00. Here is the schedule of talks.

We hope to see you there. If you can’t make it drop me a line and I can arrange to come and talk to you and your co-workers about it.
Dave

Evening Networking Event event in just 2 days!

Hi Folks,
This Thursday night, 23-Jul-2009 5:30 PM EDT - 9:00 PM EDT, I'll be participating in a panel discussion on the subject of the viability of selling SOA to the business. Other esteemed panelists include Anne Thomas Manes, Dana Gardner, Brenda Michaelson, Sandy Rogers, and of course the Zapthink Boys Ron and Jason. If you're in the Boston area I hope to see you there. Here's the registration link . Hint, use the secret code "ZAPDISC" to get a 50% discount on the event.

Details below:

ZapForum Boston: "SOA: The Night of the Living Dead" Evening Networking Event
23-Jul-2009 5:30 PM EDT - 9:00 PM EDT
M.J. O'Connor's Irish Pub
27 Columbus Avenue
Boston, MA

Dave

Today is the day we officially launch Oracle Fusion Middleware 11g.

Fusion Middleware 11gR1 is the result of a herculean effort that is 3+ years in the making.

The major areas of investment have been –
- The completion of the integration between Oracle and BEA products into unified suites. This matches to the month the schedule that we had committed to publicly when announcing the BEA Strategy last June. This continues our excellent track record of buying best-of-breed software and integrating it together into a common environment.
- Improving the efficiency of modern data centers by extending the capabilities of Application Grids. We now take advantage of new hardware and software advancements such as multi-core processors, 64-bit addressable memory, RAM-based storage, 10GB Ethernet systems, and virtualization to allow large sets of compute capacity and memory to be pooled together into virtualized grids or “clouds” that are lower cost, easier to manage with more flexible capacity to respond to business needs.
- Providing new Identity Management and Security technology to consolidate how users, their identities, and entitlements are managed, audited, and controlled to lower costs and improve security on Application Grids.
- A unified and declarative toolset with which Business Users and developers can work together to develop Business Applications & capture the behavior of the applications in metadata.
- A unified Business Process Platform with which to orchestrate humans, applications, and information into processes that can be monitored and optimized in real time, and providing a common Enterprise Portal through which people can find the Enterprise resources they need, to share them with others through personal productivity and social computing tools

Even though we are integrated into a set of suites, we are also still committed to being hot-pluggable with other technologies. We continue to support multivendor environments, and have extended our open standards support with better support for WS-*; SCA, New Identity Management standards; WSRP 2.0 Support; BPMN (Business Process Modeling Notation) and BPEL4People to name a few. In addition we expand our definition of hot-pluggable to now exploit new IT infrastructure trends that are mainstreaming by integrating FMW with these new technologies. For example we have created a new facility to build and deploy virtual appliances allowing organization to exploit Virtualization more effectively.

For more information, go here
Cheers,
Dave

Hey all!
Mark your calendar and plan to attend the Oracle Fusion Middleware 11g launch event in Washington, D.C.,on July 1. Join Oracle President Charles Phillips and Oracle Senior Vice President Thomas Kurian to see Oracle's Fusion Middleware in action. Click here to register.

If you're outside of the US, and you can't make it to Washington, D.C., then join live events that are happening around the world. Similar events are scheduled for London (July 2), Munich (July 2), Paris (July 2), Sydney (July 9), Beijing (July 14), Seoul (July 15), Tokyo (July 17), Sao Paulo (July 21), Mexico City (July 23) More Info here.

Oracle is announcing the call for submissions for the 2009 Oracle Excellence Awards for application grid solutions. This award is designed to recognize organizations using Oracle Fusion Middleware technologies to create an application grid foundation.

Does your company use Oracle Fusion Middleware application grid products such as Oracle WebLogic Server, Oracle Tuxedo, Oracle Coherence, and/or Oracle JRockit?
Nominate your organization today for a chance to be recognized for your cutting-edge grid-based solution!

An application grid enables new abilities in application infrastructure such as pooling resources, dynamically scaling capacity across clusters, caching data in memory, and automating management. If you are using Oracle WebLogic Server, Oracle Tuxedo, Oracle Coherence, or Oracle JRockit in innovative, grid-like ways, we want to hear from you!

Its easy to register. The final date for nominations is Friday, August 14, 2009

Best of luck and we look forward to hearing from you.
Dave

Come check out my talk at JavaOne this week on Wed at 11:05. Its entitled - "XTP: Patterns for Scaling SOA, WOA, and REST Predictably with a Java(TM) Technology-Based Data Grid (TS-5154)"

Here is an excerpt from the session abstract -

This session highlights specific patterns that take advantage of distributed Java™ platform agent-based caching in an in-memory data and execution grid to enable shared state management with near-in-memory access speeds for state data by services in SOA, WOA, and RESTful architectures. Using these patterns, SOA-based applications can achieve predictable scalability and high availability while insulating organizations from the need to enforce special architectural practices across the organization for "stateless" service development, enabling Java technology-based or .NET services to be written like everyday! objects that encapsulate state data with the business logic that operates on it.

The presentation discusses pros and cons of stateless versus stateful services and the service state repository. It explores architectural patterns for service state management such as fault-tolerent collection,load-balanced fault-tolerant services, business logic affinity, level 2 caching, state-based notification, and claim check. In addition, it examines pros and cons of multilevel service state caching in virtualized environments.

Come learn how next-generation SOA-based application architectures can be built to take advantage of scalable, predictable, virtualized environments that are capable of adapting to the ever-changing needs of the business.

Hope to see you there! Its in Esplanade 304-306, Moscone
Dave

An article that I co-authored, which highlites the use of large XML documents in a SOA using Application Grid,has just been published online in SOA Journal.

The article includes a downloadable example that shows how to use JAXB and a streaming STAX parser to break up large XML documents and store them in an application grid as java objects, use a distributed query to manipulate the objects in the grid, then rematerialize the data as XML upon completion of multiple grid based operations. The pattern is based on common customer scenarios we are seeing.

The article also talks about how a SOA based application would interface with an application grid by implementing a Claim Check pattern.

I wouild love to hear your comments about this subject.
Cheers,
Dave