SOA Center Service Oriented Architecture Blueprints

SOA Software continues to buy companies.

I believe that this is in a way similar to SOA Software’s buy of Blue Titan. For that they picked up Frank Martinez and some good customers, Pfizer and some others. For this, they pick up Brent Carlson, another good thinker and long time SOA industry veteran.

LogicLibrary has always fallen into the same category as Flashline (BEA acquisition and now Oracle), which is to say a repository vendor. The SOA game is a little about registry, a little about repository and a lot about policy and governance.

Anyhow, I’ve got a double Macchiato but this news isn’t waking me up this morning. Maybe AmberPoint will be acquired soon, that would probably keep me up. I’ll be in New York on Wednesday for our SOA Governance Summit. See you there.

This is pretty much a word-for-word transcript of an interview I did with IT-Business Edge about Web Oriented Architecture (WOA) and Service Oriented Architecture

http://www.itbusinessedge.com/item/?ci=41736

It reads a bit eccentrically because it’s a literal transcript, but if you know the SOA world pretty well, it gets my point across about what a Business Service is and how it transcends components, processes and event based architectural patterns.

With all the debate in the Blogosphere about WOA (like Tony Baer’s post), it’s important to answer two questions: Why are we doing SOA and How far does SOA go?

Why are we doing SOA?

I have been looking at a lot of SOA projects worldwide, and I have come to the conclusion that there is a very specific disease of Enterprise IT that SOA cures. It cures chronic ESSD, or Enterprise Software Stupidity Disease.

Diagnosing Enterprise Software Stupidity

Patient is a decades-old multi-business unit enterprise. Patient medical history includes chronic use of project-funded IT. Blood tests indicate high cholesterol and athlerosclerotic vascular system including brittle point-to-point integration and tightly coupled IT systems. Patient was recently admitted for a myocardial infarction which brought down mission critical systems and was put on a Software as a Service bypass. Business users are experiencing abdominal pain, nausea and vomiting. IT projects appear to be constipated, possibly due to a bowel obstruction. Patent is currently on WOA to reduce fever and swelling and improve short term business results.

Patient has a 15 year history of “waxy buildup” of complexity in IT that comes from layers upon layers of historical computing infrastructure and silos within silos of politically generated redundancies as well as a mish-mash of architectural styles and heterogenaety. Prognosis is that the patient requires emergency medical intervention and will be in and out of hospital care for the rest of of their lives unless put on an Architectural treatment program with regular Governance.

The Cure

SOA can certainly help with this problem, but without certain kinds of awareness, you wont get too far. Firstly it’s important to understand why you’re here in the first place. Silos dont go away, and neither to legacy systems. Because of this, you need deliberate mechanisms for dealing with both of these complexity elements.

Since my post “SOA is Over the Sky is Falling”, I hear there’s rumors on the Internets that the chattering classes are getting impatient with limited SOA success stories. The key example is Analyst and Uber Thought Leader Anne Thomas Manes who recently noted in her blog:

It has become clear to me that SOA is not working in most organizations.

Blogger Illuminatus Joe McKendrick responded with the following interpretation:

The issue, Anne points out, appears to be that SOA is still siloed within the IT organization, and the transformative effects have not yet spread to the business. Line of business managers, she observes, are not ready to share services, the crux of SOA value. Anne says so far, she has only come across one company that has been able to connect SOA success to business success, through the establishment of “strong positive and negative incentives that encourage people to adopt a better attitude toward sharing.”

I happen to agree that SOA is currently overly Siloed in IT, and there have to be ways to get the value expressed in business. Our own Mike Lees posts eloquently on the role of BPM in this.

In introducing our SOA Governance Summit event in NYC I mentioned that we have shifted from the 1% inspiration phase (Architecture) into the 99% perspiration phase (Implementation). SOA is a big and deep shift in IT.

Recently there have been a spate of posts supporting a WOA approach (including my own perspective on WOA).

* StrikeIron CEO and ZapThinker David Linthicum
* Analyst Dana Gardner
* Advisor and Consultant Dion Hinchcliffe

Now I believe that the WOA movement makes sense. But I’d like to contextualize WOA in the larger story I like to think of as Big SOA.

To be concise, Big SOA is really all of the following:

1. Component Patterns (SOA)
2. Process Patterns (POA)
3. Human Patterns (WOA)

1. SOA (the component pattern)

How can I make an argument that WOA is SOA! Well, if it’s done well in the context of Enterprise Software, WOA is a consumption pattern for SOA services. A lot of folks are stuck in thinking that SOA = Component Pattern, aka LegoLand, and this can be compounded by Web Services centrism. While components are a very powerful way to manage cost and risk in IT and to portfolio manage IT sourcing by commoditizing interfaces, if you are stuck in the Component Pattern, you will have a hard time justifying SOA to anyone but IT. Most of the business rationale in this pattern is Cost and Risk reduction. Your SOA funding ally here is the CFO and the CIO-who-reports-to-the-CFO.

2. POA (the process pattern)

Lots of folks have written successfully on Business IT alignment, and many of these folks seem to understand the Process Pattern which is predominant in business. This is the effort to create a common language between business and IT. The folks in this area are all about BPMS, BPEL, BRMS (Business Rules), and BPMN. Your SOA funding ally here is the Business. I know it’s a bit of a cheap shot, but we could call this POA or Process Oriented Architecture.

3. WOA (the human pattern)

There’s already been one counterpoint on whether WOA is all that new from SearchSOA Editor Mike Meehan The WOA crowd fit in to the Human Patterns area. When I say Human Patterns, the top side of it includes new user interface technologies (RIA, AJAX, etc), the Internet, Saas, Cloud Computing, Mashups, Semantics, Syndication (RSS, ATOM, etc), MetaAggregation (RottenTomatoes, MetaFilter, MetaCritic, etc.), Social Networks (Open Social, Google Social Graph API, FBML), Social Bookmarking (Digg, Del.icio.us, StumbleUpon), Web 2.0 etc. Much of the programmatic aspect of Human Patterns more closely resemble Event Driven architecture (EDA), Exception Management, Complex Event Processing (CEP), Rule Engines, Publish/Subscribe, and to some extent IaaS (Information as a Service).

This is the huge and frothy ocean of the Internet. The Internet has recently become more China centric than US centric. As a market mechanism, the Internet is increasingly a “Force Multiplier”. The collaborative filtering power, recommendations and social graph event propagation (e.g.Twitter) enable “Flash Commerce” (word-of-mouth smash hit indie movies like Juno), Long-Tail Commerce and Super Product Commerce (Based on the Wii, Nintendo bigger market cap than Sony in Japan). But what’s truly emerging is the notion that the Internet has “physics” including a momentum term. Just look at the distribution of Facebook Applications.

Leveraging the kind of competitive differentiation power of the Internet extends far beyond just brand space and into how organizations can provide differentiated products and services.

If you look at the iPhone, you could see it as a “composite app” of a bunch of “old” technologies (BSD based Mac OSX, Commodity Flash memory, Wifi ASICs, etc.) “mashed up” with the multi-touch screen. The ability to execute business in this environment has to do with creating excitement on the Internet (sure Apple didn’t spend a lot of conventional marketing dollars on iPhone) but also to do with managing supply chains and ensuring that you can meet the demand.

WOA combines the interest of young Web 2.0 hackers with innovative Marketing people. As consumption patterns for SOA, WOA has the potential to drive competitive differentiation and provide new consumption patterns for core Enterprise services. WOA also tends to be internally driven by younger Internet-age workers who just want to crank out Ruby and hack Web 2.0. Getting exciting innovation out of WOA is quicker, cheaper and easier than working the Componentization and Governance of IT infrastructure, that’s for sure. So the funding allies for WOA include “guerilla funding” (no official funding) as well as Marketing people.

Since SOA is Seismological, consider that it is an inevitability, but only if you look at it in the geological time scale. WOA is sociological and therefore has a greater chance of establishing momentum in the short term. What remains to be seen is if WOA, SOA and POA emerge as a postmodern rethinking of the role of technology in business.

Conclusion SOA + POA + WOA = SOA

Not everyone involved with SOA would agree with this equation. Some would say legitimately that WOA has nothing to do with SOA.

I think these are legitimate views.

But since SOA is generally accepted to be a paradigm or style of thinking, I would like to encourage us to think in expansive terms of what “service” means. To me, service involves those who serve (”providers”) and those who are served (”consumers”). But it also encompasses the very human elements of how we feel when we are well-served and the reciprocal loyalty between those who serve well and those who are well served.

To compete in business, it is important to both serve well and to be well-served.

There’s a lot of talk about SOA and WOA led by ZDNet Blogger, Analyst and great guy Dana Gardner.

WOA is a very fast paced “Web 2.0-ish” way to innovate and mash up concepts quickly. And the massive power of it is that you are standing on the shoulders of Giants who have built “Social Utilities” and amazing Web-Oriented APIs so that you can use astonishingly simple expressions and get vasty goodness (technical term) very easily.

The article states…

WOA has evolved via massive scale trial-and-error, and so has been designed through viral adoption, user pull, self-help and with self-qualification of real-time productivity in mind. It works because it just works, not because it’s supposed to work, or because someday it will work. As Dion says, “And these new models intrinsically take advantage of the important properties of the Web that have made it the most successful network in history.”

It really is a phenomenal quick start to value and a great way to interconnect people to one another. This is a way to extend brand, compete, offer new applications, listen to events on the network, offer compelling new user experiences, participate in “long tail” economics, and just plain try out a million things cheaply.

In fact, the key idea of letting a thousand or million flowers bloom on the “platform” and let nature select the winners was the theme of my SOA World Keynote.

The interesting challenge for the Enterprise is how can the core components that generate the majority of revenue for the enterprise be successfully deployed as a platform? It’s one thing to take advantage of everyone else’s platform out on the internet and it’s a completely different thing to offer platform services of your own.

The experience of being a consumer of services is very snappy and cool and you can use this WOA experience to show value quickly and mash up some amazing things. You can leverage the massive power (and in many cases free hosting, viral application distribution, and software services) of Web Oriented platforms that are offered and for the most part freely available to anyone with markup skills and maybe a scripting language like Ruby Python PHP and JavaScript.

So go kick some butt on WOA. It will do you a world of good. But keep in mind that having something of value to offer the Internet and distributing those services (whether freely or not) enables you to add value to your customers and your entire value ecosystem. By offering your own business services on the network, you can encourage all of the innovation in your ecosystem to build on top of *your* enterprise platform in a wild and open way… and let the laws of nature select which ones will be successful for you.

This will give you the “Peacock’s tail”, which is sort of a tail made up of a lot of tail feathers. You’ll find that having many many customers and many many different ways to serve them is going to make your offerings compelling, sticky and value adding.

My 2 cents,

Miko

I’ve been thinking about the drivers for SOA for a while now and think that there are two major directions initially taken up the SOA Mountain. Regardless of which path you travel you end up on the top of the mountain (assuming you implement well).

This is based on the idea that IT by itself isnt funded to do big bang SOA. So you need to make alliances. The first kind of alliance is centered around the notion of “Sustainability”. Why is this an important theme? Because it aligns the IT organization with the CFO organization. The CFO organization looks on the budget spreadsheet for IT with lots of distaste. There are large line items traceable to IT that pay for necessary evils and emergencies. Huge security breach? Huge regulatory compliance cost? Heavy software maintainance fees? Expensive software integration projects? Why is software not done in a sustainable way? too many unmaintainable, incompatible, insecure blobs of code.

So the idea of achieving Lifetime Total Cost of Ownership (TCO) or a Return on Asset (ROA) model for IT services is an attractive one to the CFO.

The other road up the mountain is Global Optimization. This can create alignment between the business function and the IT function. So instead of a disconnected Business Process team drawing process diagrams on a white board, there’s a smooth connection between the BPM world of the business and the SOA world of IT. Instead of “good for me, bad for you”, there’s a new paradigm in town to make things optimal for everyone.

The drivers in this case are business agility and rapid response via component assembly. Businesses tend to be focused on business processes, so the connection between BPM and SOA in this path tend to be significant.

So in short, IT cant achieve SOA by itself, and needs to focus on cleaning up cost themes for the CFO, or alternatively can focus on speeding up avenues to revenue for the business via BPM.

I’ve been hearing a lot about SOA Governance is something you do not something you buy.

Dave Linthicum says it. Anne Thomas Manes says it.

I happen to agree… but I want to make sure we dont overcorrect. I responded to Dave’s post with the following:

Sculpting is also something you do, not something you buy.

Buying a chisel doesnt make you Michaelangelo

However, even Michaelangelo doesnt try to carve marble with his bare hands.

So get the best people AND the best tools and you’ll have the best outcome for SOA.

Philip Howard’s piece titled Merchant relational databases: over-engineered and out-of-date? is a fascinating look at some emerging trends in Database technology.

The research piece references a recently published paper by Michael Stonebraker et al.

It states:

We conclude that the current RDBMS code lines, while
attempting to be a “one size fits all” solution, in fact, excel at
nothing. Hence, they are 25 year old legacy code lines that should
be retired in favor of a collection of “from scratch” specialized
engines. The DBMS vendors (and the research community)
should start with a clean sheet of paper and design systems for
tomorrow’s requirements, not continue to push code lines and
architectures designed for yesterday’s needs.

Now obviously everyone knows Stonebraker and expects him to be shaking things up. He’s behind Streambase, the CEP company and of course founded Ingres, Illustra, Cohera, StreamBase Systems and Vertica and was previously the CTO of Informix.

This does not only apply to CEP formulations, where it’s clear that even existing vendors have to scramble to move towards Hierarchical (rather than flat) data models. But it applies to OLTP and every other database arena…

We live in exciting times…

I understand the reaction people have to the phrase SOA Governance.

Mike Meehan of SearchWebServices says:

…let’s be honest, the term “SOA governance” sucks. It reeks of someone else telling you what to do, hectoring you over every little detail of a project. It sounds about as desirable as a colonoscopy with an IMAX camera.

Still, I’ve wound my way around and around this thing and keep coming back to it.

We have a phrase at Software AG webMethods which is “Liberate and Govern”, which are two sides of the same coin. Perhaps “regulate” might be a better word, but if we’re planning to ditch “govern” we had better move to something a lot better. Regulate doesnt really get me excited either.

The interdependencies in SOA drive complex systems and therefore emergent behavior. Governance as a term fails to capture the essence of the notion of complex behaviors made emergent by the use of simple rules of engagement. I’ve been favoring the word “Coordination” lately, which evokes a flock of geese or ordered “swarm” behaviors.

So I think we are stuck with the term, but it’s important to recognize that we are talking about the emergent behavior of complex systems. If we want to consider the entirety of the system which should include the Service Pattern, the Process Pattern and the Event/Exception/User Pattern (most broad definitions of SOA do), then we need to understand that SOA Governance is a fully human cybernetic system and therefore subject to dynamic systems principles.

The Tao Te Ching chapter 59 is clear on this subject:

In caring for others and serving heaven, There is nothing like using restraint.
Restraint depends on giving up one’s own ideas.
This depends on Virtue gathered in the past.
If there is a good store of Virtue, then nothing is impossible.
If nothing is impossible, then there are no limits.
If a man knows no limits, then he is fit to be a ruler.
The mother principle of ruling holds good for a long time.
This is called having deep roots and a firm foundation, the Tao of long life and eternal vision.

Another translation of the same verse

Chapter 59

In governing people and serving Heaven
There is nothing like conservation
Only with conservation is it called submitting early
Submitting early is called emphasis on accumulating virtues
Accumulating virtues means there is nothing one cannot overcome
When there is nothing that one cannot overcome
One’s limits are unknown
The limitations being unknown, one can possess sovereignty
With this mother principle of power, one can be everlasting
This is called deep roots and firm foundation
The Tao of longevity and lasting vision

The second translation uses the phrase “governing people” whereas the first says “caring for others and serving heaven”. The goal in both is stated to be the establishment of deep roots and a firm foundation for longevity. That’s really at the heart of SOA and SOA Governance.

My 2 cents,
Miko

Game on microsoft!

Conference Call Details Here

Microsoft’s PowerPoint preso about it is here

The four themes are:
Expanded R&D capacity
Emerging User Experiences
Scale Economics
and
Operational Efficiencies

I’ll talk most about Emerging User Experiences, especially in Mobile

Microsoft takes the fight to google with unsolicited 45 billion dollar acquisition offer.

62pct premium over street price. Damn! Henry Blodget has the financials.

Unsolicited offer–the culmination of lying in wait for several years, plus the dip in Yahoo! stock price and the announcement of laying of 1000 employees. Interestingly, Yahoo has been quietly building office space in Bellvue Washington for another 3,500 employees. Ballmer clearly taking advantage of the occasion to buy at a better price.

Terry Semel gets ousted the same day as Chairman. dirty details here.

Big implications for the consumerization of the enterprise and of course the future of the compute cloud.

Some of the properties like Flikr are best in category and I believe Yahoo and Microsofts dual focus on the mobile market will produce a huge comptitor to google/oha android.

Although Eric Schmidt is recusing himself from Apple iPhone discussions on ther board meetings, the handset OS world will probably still include MacOS (!) But Linux based OS like Nokia Trolltech QT and Google/OHA android will compete with Windows Mobile for the lions share.

Interesting to see that the thee giants of handset OS are Windows, Linux and MacOSX.

The mobile opportunity is the massive one, coming out of high growth emerging markets like BRIC (Brazil Russia India China). China Mobile alone is more than a third of a billion users.

No wonder the CEO of china mobile is sharing the stage with the ceo of google at the WEF in Davos.

Watch for the clash of the titans (GOOG v MSFT) to happen in the mobile space.

Google is bidding on the 700Mhz spectrum and launching Android, but Microsoft has any entrenched mobile OS and some surprising novel services to add coming from Yahoo! (Like Flikr)