November 19th, 2008
SOA: it’s only rocket science
What can go wrong with SOA? Let me count the ways.
Better yet, Miko Matsumura, Bjoern Brauel, and Jignesh Shah, all with SoftwareAG, have done the counting for us, in a new book that shows how to overcome various SOA issues, called “SOA Adoption for Dummies.” (Available for download here.)
Here are some of the issues that will torpedo an SOA initiative:
- SOA fatigue: “The duration and sheer complexity of SOA can lead to organizational fatigue.” The authors say this can be minimized by ensuring that each project continues to show a return on investment (ROI).
- Leadership changes: “Key individuals can get fired, get a better job somewhere else,
or retire. Anything can happen.” The authors say that such sudden changes can be overcome with a shared vision and committed team. - Vendor backlash: “Software vendors may promote their stack as a single-vendor SOA solution. A vendor may find a way to tightly couple or avoid interoperability in an effort to maximize its revenue from software licenses or services.”
- Consultant backlash: “Consultants want to stay in control and sell more billable hours. SOA can either benefit or jeopardize this lucrative agenda. Be aware that consultants will fight to defend or expand their revenue in most cases.”
- Implementer backlash: “Unless policy enforcement is gradually phased in alongside motivational incentives, developers and other key stakeholders may actively or
passively resist.” - Funding cutoff: “Unless executives continue seeing measurable business results coming from SOA projects, there may be a cutoff of funds going to your SOA program,” the authors says. The urge keeping an eye on the measurements that “matter most to your executives and make sure your SOA program aligns with those objectives.”
The authors say the best tool for managing these threats is the use of continuous
measurement — not just IT metrics but business metrics. “By managing expectations and showing the positive business results that come from SOA, you can keep implementers, funders, and other key stakeholders in your organization enthusiastic about SOA.”
Who pays for SOA? That’s a good question, because SOA is a lot more than a tangible product or offering:
“You can usually explain much more easily the value of something you buy versus the value of something you do. Because SOA is something you ‘do,’ not something you buy, the trick is figuring out how to get your organization to fully embrace it.”
The authors do a good job of laying out the fundamentals of SOA, and the book is loaded with helpful tips. For example, the authors recommend employing business process management as a way to quickly link SOA to business processes:
“Coupling SOA with BPM ensures consistency without hard wiring of code. When you develop processes in a BPM environment, you quickly realize which services the SOA platform has to provide in order for processes to be automated. BPM gives scope to an SOA initiative in such a way that SOA is seen immediately as adding value versus adding cost and complexity. Because SOA is seen from the perspective of the business, the SOA infrastructure serves its intended purpose: to provide valuable service, not just some form of interface.”
The authors also equate the SOA journey to a space flight, as follows:
“SOA adoption, like a real-world rocket,experiences a danger zone between blast-off and the weightlessness of orbit. When fully realized, SOA can transform your business. But until firmly established, your SOA dreams can plummet back to earth. Getting across this SOA danger zone requires a focus on several key principles: Keep the pointy end of the SOA rocket up by measuring your progress and making course corrections as you go; Keep moving up by motivating the teams and players in your SOA adoption; Don’t stop till you’re weightless by automating processes until implementing SOA becomes second nature and therefore effortless.”
Fittingly, the last chapter, entitled “To Infinity and Beyond,” looks at where SOA is leading, and points to new worlds such as event driven architecture (EDA) and complex event processing (CEP), and software as a service and platform as a service,
November 17th, 2008
Deep down inside, on a subconscious level, the business loves SOA, right?
I recall the words of one observer a couple of years back: “How many businesspeople do you know have ever come begging to you for SOA?” Geek & Poke’s Oliver Widder picked up on my latest post on SOA business issues behind SOA acceptance:
November 15th, 2008
IDC-HP study: SOA targets the business, but is the business sold on SOA?
SOA success may be in the eye of the beholder — and that is a factor holding back SOA-based initiatives, Sandy Rogers, analyst with IDC, writes in a new study, underwritten by HP. As Sandy put it: “Individuals shepherding their organizations’ SOA ventures… reported that SOA still is most often perceived by others in their organizations as a technical initiative, especially at the onset.”
CIOs, developers, operations, and business see SOA through different lenses
However, make no mistake about it, SOA proponents see service oriented architecture as a business endeavor, she added. “Influencers understand that overall SOA success requires better business alignment, shared consensus, and deeper levels of participation and as such are particularly focused on driving those dynamics.”
IDC’s study of 109 companies found that 40% of organizations had SOA efforts underway, with another 30% in the planning and pilot stages. This tracks very closely with other surveys from analyst groups, which also find about 70% to 80% of companies are employing SOA methodologies or are considering doing so. Some interpreted lack of growth in planned projects as SOA support wavering, but I see it more as close to saturation point.
SOA may have reached as many organizations as it will likely reach, but is maturing as the preferred approach to enterprise systems development and deployment. In the IDC report, however, Sandy indicates that SOA is now penetrating deeper into the enterprise. As Sandy told ZDNet colleague Dana Gardner in a recent podcast, “As we go to this generation, SOA, in and of itself, is spawning the ability to address new types of models, such as event-based processing, model-based processing, cloud computing, and appliances. We’re really, as a foundation, looking to make a strategic move. …if [organizations] haven’t already adopted SOA, they are planning on it, and at greater levels of engagement. So, if it is not going to be ‘the’ standard for most or all systems, it’s important, and will be used for all new projects, or it’s a preferred approach.”
However, as outlined in the report, Sandy noted that different stakeholders have different perspectives on SOA, and thus, may work at cross-purposes:
“The specific goals, experiences, and perspectives of a CIO may be somewhat different from those of a chief architect and then again of someone in the development ranks or on a quality team,” she said. Chief architects are often most focused on establishing an SOA program and expanding its pervasive use…. Yet, for a developer, SOA success may mean something different, perhaps the ability to achieve one’s project on time or how easy it is to attend to one’s tasks on a day-to-day basis…. Operations staff may be more concerned about guaranteeing service security, availability, and reliability, along with managing the impact on underlying systems.”
Of course, there’s ultimately the business perspective, which addresses capabilities such as more access to information, faster time to market, more collaboration, and more effective integration.
Another issue that hampers SOA-based initiatives is lack of effective governance — early on. “Nearly all those interviewed in our study reported that if they could do one thing differently, they would have started a stronger and more automated SOA governance program earlier,” Sandy wrote in the report. “A few participants highlighted that time spent debating and evaluating SOA technologies could have been much better used in defining standards, policies, and best practices, and then subsequently ensuring that any infrastructure selected would accommodate these concerns.”
Interestingly, some SOA-based efforts may even be too successful early on as well. “Some challenges may arise when attempting to scale out and build greater participation,” Sandy said. “Greater coordination between service providers and consumers needs to be nurtured. Many organizations in our study become, as one interviewee noted, ‘a victim of success,’ as key stakeholders start to ramp up service development and use before any real foundation is laid.”
Sandy also hit upon a point often raised at this blogsite, and that is the inherent challenges in measuring the success of an SOA effort. Beyond the initial and more apparent successes with IT and development streamlining, how much business success can be attributed to the SOA initiative?
“According to IDC research, many IT professionals embarking on SOA programs typically envision their primary benefit to be code and service reuse, whereas those having already implemented SOA to some degree often report the most important capabilities obtained as overall flexibility and ability to respond with solutions faster to market. It is interesting that the former objective is more tactical and technically focused, and the latter more about strategic alignment and support for overall business pursuits. Therefore, approaches taken while adopting SOA may vary and be more or less constrained by individual perceptions. SOA may also be measured along a potentially diverse span of expectations. Nothing is assured, however.”
The takeaway here is that SOA is now well-established as a methodology and practice across a majority of organizations. The key now is to advance SOA practices from that of IT optimization to more of a business transformation initiative.
November 14th, 2008
Whatever happened to the ‘middle’ in middleware?
Quote of the week:
“In the mid 1990s, ‘middle’ meant ‘the gaps between applications and software components.’ Middleware was technology you turned to in order to try to build distributed systems: transaction processing middleware, database middleware, object middleware… Middleware was a technological concept. …Now, ‘middle’ means ‘the gap in a technology stack between an operating system and packaged applications.’ Middleware is now defined largely by vendors from a software product marketing perspective, rather than by customers from a technology perspective. Consider the portfolios of IBM, TIBCO, SAP, Oracle: they all talk about ‘middleware stacks,’ but these things include process management, content management, master data management, and business intelligence tools - and sometimes even DBMSs.”
November 13th, 2008
SOA’s ‘decline’ may be overstated, but companies still want results
Fred Cummins over at EDS looked at Gartner’s recent downbeat survey (also reported here), which stated that fewer companies are launching SOA, and speculates that the economy may be putting the kibosh on new initiatives. The Gartner survey was conducted in early summer. “I suggest the primary driver was the declining economy, and that if another survey were to be conducted today, the decline in SOA adoption plans would be more severe.”
SOA is here to stay, but needs more business-sense
By the way, in my own work with Evans Data, based on a survey conducted in September and October (when the financial crisis was hot and heavy), we that found all systems were still ‘go’ for SOA efforts — but with plenty of continuing reservations about ROI. Fifty-two percent had SOA-based projects underway, and another 31% were planning to start SOA projects over the next 12 months. Gartner put it at 53% with SOA and 25% planning earlier in the summer. So, if anything, 78%-84% of organizations with or planning SOA seems more like saturation than anything else.
ZDNet colleague Dana Gartner also just spoke with IDC’s Sandy Rogers, who finds in her polling that SOA adoption has deepened across many enterprises. “The issue is not necessarily deciding if they should go toward SOA. What we’re finding is that for most organizations this is the way that they are going to move, and the question is just navigating how to best do that for the best value and for better success,” according to Sandy.
Nevertheless, Fred Cummins’ advice on the direction SOA needs to take in light of economic stress makes plenty of sense. Target SOA-based projects at streamlining and efficiency. “The business imperative should be consolidation,” he says. “That’s what businesses do when they have to cut back. SOA should be viewed as the architecture for consolidation-for enabling economies of scale in capabilities that can be shared across lines of business. The design of these consolidated services should reflect a long-term, SOA strategy where shared services become fundamental to the design of the enterprise.”
But, as analysts are suggesting, are organizations really feeling disillusioned about SOA, or is it simply too early to tell? Are there actually a lot of emerging success stories underneath the radar? If SOA is a tougher sell, what can SOA proponents do to increase enterprise acceptance? To answer these questions and more, I will be moderating a panel discussion as part of next week’s SOA in Action virtual conference over at ebizQ. In a discussion on “Avoiding SOA Disillusionment,” I will be joined by industry luminaries including Phil Wainewright (a familiar face here at ZDNet), Dr. Chris Harding (The Open Group), Jignesh Shah (Software AG), David Bressler (Progress Software), and John Michelsen (iTKO).
Issues to be tackled include SOA complexity, virtualization/availability of services, architecture approach, trust issues, organizational issues, and complementary approaches such as cloud computing and SaaS. I’ll also raise Dan Woods’ point that perhaps SOA shifts too much to already overworked IT departments.
November 11th, 2008
Forbes exposes SOA’s ‘dirty little secret’
In his latest post in Forbes, Dan Woods delivers an interesting revelation: with SOA, IT departments end up taking on a lot of the work formerly handled by software vendors.
IT departments as software manufacturers
In the old days, a company took an application package in and ran it. (Never mind all the integration work and training and teeth-gnashing, but that’s another story.) With SOA, systems and applications are decomposed and re-assembled as services. These reconstituted, re-assembled apps, of course, belong to the assembler, likely the IT department. “For a mash-up or composite application, there is no manufacturer,” Woods says. “The IT department or the person who assembled the application plays that role.” And with that role comes the need for operational and business process monitoring.
So, rather than withering away, as predicted by some, we may need even more IT. As Woods puts it:
“The challenge of full-blown SOA means that IT departments have many of the same responsibilities of software manufacturers and a few new ones. SOA management vendors like Hewlett-Packard, AmberPoint and Mashery are putting a lot of work into products that help with governance and monitoring. Software manufacturers like SAP are building consistent collections of services delivered as products. …the mix of services needed at any one company will likely come from many sources, which means a whole new set of skills will be required for IT departments.”
Woods says this may be all too much for already overburdened IT departments, so some of this work will be taken on by a new breed of service aggregators, who will “step in and collect large amounts of services, solve the integration problems and guarantee quality of service and scalability.” But the cavalry probably won’t arrive in force for some time, so the work is still left to IT.
Agree or disagree? It would be interesting to see some study into this area — is there a correlation between SOA efforts and IT staff sizes, controlling for the fact that larger companies with large IT staffs tend to be further along with SOA?
I think even with packaged applications, there has always been a lot of additional work (and headaches) for IT, from installation to user frustration to performance issues to integration with other applications and systems within the enterprise. And there’s still coding required. If anything, SOA alleviates the burden of writing a new interface every time there’s a need to leverage these packaged applications. Plus, SOA reduces the risk of breaking applications, since most of the change work takes place in the service or interface.
Woods adds mashups to the SOA scenario, so it should be noted that another trend that is taking shape is that end users will increasingly take responsibility for their own front-end applications and interfaces.
Along with user-generated applications, we’re seeing more automation. Here’s something for the future — we’ll someday maybe see the reality of the self-assembled application or service. Systems will predictively determine emerging requirements and automatically and dynamically generate mashups or composite applications that align with the new demands.
But there will always be enough going on to keep IT busy.
November 10th, 2008
Still coding after all these years….
SOA nirvana may still be a ways off…
Geek & Poke’s Oliver Widder picked up on my recent post on Gartner’s downbeat SOA assessment, and notes that with the glacial pace of progress, hard-coding will probably be with us for some time to come.
November 7th, 2008
If SOA were an issue in this week’s presidential election…
Quote of the week:
(Great hypothetical of a presidential candidate’s stand on REST…)
“So the good folks of America need to know more about this so-called REST architectural style. Who is REST, and why does he want to hurt Joe the Systems Integrator? REST sounds dangerous. Joe the System Integrator realizes that this is not the time to constrain the Big Vendors Integrators Dream - Joe knows that it’s patriotic duty to spread the complexity tax from the top down. My friends, the big vendors have committed to a 10% reduction of SOA spread over the next 10 years - and we issue vouchers for you to claim back your Advanced WSDL 2 certificate costs as a tax break. REST will just raise your complexity and kick your dog.”
-David Ing, From 9 till 2 blog
November 6th, 2008
Gartner: SOA sinking into trough of disillusionment
Did Gartner find SOA has entered its so-called “trough of disillusionment”?
A new survey by the consultancy found that SOA doesn’t seem to be as popular as it was a year ago. Why? Is it the economy? Maybe — but it may be a maturing of the methodology, tempered by the recognition that SOA won’t deliver results overnight.
A new report out of analyst firm Gartner says that since the beginning of 2008, “there has been a dramatic fall in the number of organizations that are planning to adopt SOA for the first time.” Last year, in 2007, 53% of organizations in surveys were planning to launch their first SOA forays, a number that dropped to 25% this year.
Currently, 53% of the companies Gartner surveyed have SOA-based projects in progress. (Presumably, many would have been those in the planning stages in last year’s survey.) Sixteen percent had no plans at all to adopt SOA, up from six percent in 2007.
Daniel Sholler, research vice president at Gartner, said organizations are more aware these days that SOA is not something adopted on the fly, and that a deliberative governance process is needed to make SOA a worthy business venture. And, as also noted frequently here at this blogsite, there’s a shortage of professionals with the skills to bring it all together:
“Organizations without a clear business case for SOA and without a plan to develop or acquire the necessary skills are justified in taking a cautious approach, and delaying SOA adoption plans for the coming year. The focus should be on creating shared services and the governance processes necessary for sharing within a reasonable domain. Larger organizations (more than 5,000 employees) are challenged to create enterprise governance.”
The Gartner statement observes that “the two major reasons that organizations choose for not pursuing SOA are a lack of skills and expertise, and no viable business case.” Gartner also said that “conversations with many Gartner clients have shown that there is a great deal of confusion about how to construct a business case for SOA. Even if a valid business case exists, then the required skills are often unavailable in-house, and the costs and effort to develop in-house skills and acquire outside expertise are often daunting.”
Remember, as Gartner tells us, once a technology approach passes through the trough of disillusionment (because it didn’t live up to the hype), it enters a quieter period in which practitioners roll up their sleeves, sit down, and figure out ways to make it all work.
UPDATE: Miko Matsumura reminds us of the Nucleus Research study that found that “only” 37% of organizations are seeing ROI from their SOA efforts. Many read into those results as proof that SOA is a dud; but, frankly, 37% wasn’t too shabby of a number for something companies have just barely begun to embark on. There are strong opinions on both sides of the matter — reader bjbrock, for one, says SOA has failed to show ROI thus far.
November 6th, 2008
Survey: Companies struggle with SOA’s hard-to-pin-down ROI
Wanted: professionals and managers that not only can talk to the business, but also devise ways to build the business case that delivers measurable return on investment.
I recently helped write up the latest Evans Data survey on SOA and Web services, which found that determining the return on investment for a SOA project is the greatest challenge facing developers working on SOA programs. The survey of almost 400 developers and managers finds that almost one in five cited ROI as the most challenging aspect, more than identifying available Web services, testing and validation, or absorbing the expense.
Savings from services reuse is the most common economic benefit in SOA projects globally, the survey also found. The issue is that ROI is something that needs to be measured from business results from areas across the enterprise. As discussed frequently at this blogsite, initial results from SOA-based efforts are delivering IT cost efficiencies, but the impact on business agility or improved business processes is tough to measure.
The survey also found that four out of five participating organizations, 84%, either have service oriented architecture in place or plan to do so within the next 12 months. For the most part, Evans Data surveys going back to 2004 have found SOA to be underway or planned in a minority of companies. It was only at the end of 2007 that this number crossed in to the majority, when 60% reported having SOA in place, or planning for adoption within the coming year.
The survey also found that companies desperately need enterprise or SOA architects with both business and technical skills. Within the next 12 months, two out of five respondents will be engaged full-time in SOA and Web services projects. In addition, close to seven out of ten respondents report they need professionals skilled in the merging of business and technology, and would consider creating a special role within their organizations for an architect who can very precisely translate business processes into IT services.
The survey, conducted in September and October, revealed that those using Java for SOA projects are more likely to be creating applications for outside clients, while those using .NET tend more to be internal developers. Microsoft and IBM are the two companies that command developers’ attention when it comes to SOA and Web services.
November 5th, 2008
Conflict between Rich Internet Apps and SOA? Say it isn’t so
Okay, it’s well documented that the worlds of SOA and Web 2.0, while increasingly overlapping, have operated in parallel dimensions. SOA preaches ‘worry about the business, and the technology will follow,’ while Web 2.0 preaches, ‘get into the technology, and the business will follow.’
Still, the headline “Resolving RIA-SOA Conflict” is an eye-catcher, especially since we’re counting heavily on these two methodologies to get along and perform in unison. Rich Internet Applications are potentially part of that “last mile” between SOA-based infrastructures and client endpoints. They are the foundations of simple, lightweight front-ends and mashups that are becoming so popular, and are enabling business users to get in on the SOA action with their own rapidly deployable composite applications.
In a recent article, Michael Poulin points out that the issue between SOA and RIA is that they often differ too widely in granularity. “The major disagreement between RIA and SOA is in the fine-grained operations in RIA and the coarse-grained type of interfaces of SOA business services.”
SOA and RIA projects often see a “mismatch in objective requirements, granularity, and data formats.” That’s because the “RIA spectrum is wide,” Poulin says. RIA encompasses applications with interfaces for information reporting, modifying predefined business data, collecting and inserting new data into the systems, fast and frequent exchange of information in social/community-oriented Internet applications, and setting commands in the processing systems. “Depending on the task, RIA might require a frequent and fine-grained information exchange between an RIA client, such as a browser and a RIA application that is deployed on the server.”
Poulin proposes an “RIA-SOA collaboration design pattern” to resolve the discrepancies between RIA and SOA. This pattern involves the use of a “conciliator module” between RIA requests and SOA business service interfaces. The conciliator module “bridges business functionally provided by SOA business services and business user interface functionality,” while supporting the delivery of data to and from RIA-based widgets.
It’s urgent that any issues between RIAs (and mashups) and the SOA-based infrastructure be resolved, as this now represents the path of least resistance to service orientation.
November 2nd, 2008
Microsoft: shaky future or rocking and rolling?
Okay, I couldn’t resist plugging the analogy into the latest BriefingsDirect analyst panel discussion, but isn’t this true when it comes to Microsoft: That Microsoft’s target constituency from day one has always been the IT-technically challenged business users, a vast number of which are small business users? Is Microsoft not the IT solution for the “Joe the Plumbers” of the world?
Can Microsoft turn Joe the Plumber on to SOA?
That, I feel, has always felt was the secret of Microsoft’s success from 1980 on. And I continue to see evidence of that original intent in Microsoft’s SOA pronouncements. It’s unclear, though, how they will handle the growing storm clouds coming from, well, cloud computing.
Dana Gardner, Tony Baer, Mike Meehan, Brad Shimmin, Jim Kolebius, Dave Linthicum, and yours truly recently had an invigorating discussion on the future of Microsoft, weighing the impact of all the various pressures on the company’s margins — from cloud computing to open source to the economy in general. (Transcript available here.)
Will the software giant be even bigger in five years, or will it begin to shrink? By a narrow four-to-three margin, most of our group said Microsoft will be larger in five years than it is now, mainly because we expect the IT industry as a whole to be larger by then. The other three analysts predict Microsoft’s revenue base will erode.
Dana pointed out that Microsoft seemed to missing the boat when it came to many of the top trends shaping the IT space in 2009, as defined by Gartner: cloud computing, blade computing, Web oriented computing, mashups, specialized systems, social software, business intelligence, and green IT. The vendor is making progress in some other hot areas of IT, including virtualization and unified communications.
Dave Linthicum summed up Micrososft’s shaky future this way, observing that “Microsoft isn’t going to go away, but I think they’re going to find that their market has changed around them. The desktop isn’t as significant as it once was. People aren’t going to want to upgrade Office every year. They’re not going to want to upgrade their desktop operating systems every year. Apple Macs are making big inroads into their market space, and it’s going to be a very tough fight for them. I think they’re going to be a lot smaller company in five years than they are today.”
Mike Meehan says the tough will be going at the enterprise level as well, stating that “Microsoft has been capped on the business side.” Microsoft has lost ground in the service-oriented space as well, he feels. “My view is that .NET has lost to Java, just as an enterprise technology. It’s a niche. It’s an avenue where Microsoft is going to have a presence. People are going to use Visual Studio. They can build out Oslo and they can try to keep people in with as much service orientation as Microsoft can give you in their package, but they are not going to be on the same par as IBM, Oracle, or even SAP long term, in terms of being able to give you enterprise applications and application development tools.”
Jim Kolebius begged to differ with such dour appraisals of Microsoft’s prospects, noting that the vendor has made great strides into the enterprise with SharePoint and SQL Server. “SharePoint is everywhere,” he said. “It’s the de-facto standard portal for a large swath of the corporate world.”
My point of view is that Microsoft has been very adept at entering unserved and underserved markets — including smaller companies and departments of larger ones with small budgets, as well as non-technical business users — and continues to pursue that strategy. Survey after survey I have seen or been involved with — to this day — always put Microsoft tools, platforms, and applications on top.
Microsoft has a pretty good shot at solid growth in the coming years — as long as they avoid doing any more Jerry Seinfeld commercials, that is….
October 31st, 2008
SOA fear factor
Oftentimes, the SOA market can seem a little scary, as Greg the Architect finds out. Happy Halloween!
October 30th, 2008
Best SOA governance is hard to see
Quote of the day:
“My view on governance is that it is about guiding an organization to a desired behavior. If the organization is behaving that way, no one even realizes that governance is there, simply because the desired behavior has become second nature.”
- Todd Biske, author, SOA Governance
October 29th, 2008
Who buys the first round of SOA?
Okay, here’s an awkward issue that still hasn’t been resolved in many organizations. SOA promises a lot of great things for the enterprise, but… Who buys the first round? Someone, after all, has to put up the money to get it all started.
In a new post, EDS fellow Fred Cummins says true enterprise-focused SOA can’t doesn’t spring up out of a single silo; there needs to be some sort of SOA infrastructure in place. Imagine being the first company to install telephones, but also being expected to finance the costs of the entire network, he says.
For SOA projects, departments or business units creating and maintaining the services may end up bearing the start-up costs; with secondary participants in the SOA network sucking up the benefits without the risk.
Corporate needs to step in an lay the foundation and commit the seed funding, so all departments can partake in service orientation. However, the need for this more centralized infrastructure is seen as a “burden” on organizations — the business executives aren’t convinced.
The problem is, for these early stages, SOA is sold as an IT efficiency improver. As Cummins puts it:
“The infrastructure investment anticipates a return when the number of network participants reaches critical mass. Unfortunately, most executive leaders will be skeptical about the potential business value of SOA. SOA is promoted primarily as a new technology with potential IT cost savings.”
Not enough is being done to sell the ultimate value of SOA, which is in enabling improvements to the efficiency and agility of the enterprise, Cummins writes.
This must be demonstrated to executives, not only for the infrastructure investment, but for the changes that will be needed in organizations and governance. “The real challenge is to find redundant capabilities that can be consolidated to yield significant business value,” Cummins says. He says the business value needs to be sold from the get-go. And that business value “comes from the consolidation and sharing of business capabilities across the enterprise. There may be some business value in the consolidation of computer applications and supporting technology, but the real business value should come from transformation of the enterprise to exploit shared capabilities.”
The trouble is that many SOA efforts hit the wall when it comes to moving ROI into this realm. It’s an issue of measurement. ROI is relatively easy to capture and see when developer productivity is realized, or when data entry time is reduced, or when an interactive voice response channel can operate off the same system as the customer care center.
But business agility is squishy stuff. How do you measure SOA’s contribution to business transformation or business agility? Again, very squishy stuff.
October 28th, 2008
Situational awareness, thanks to SOA-driven dashboard
Everyone knows what a dashboard can do, and yes, that’s SOA behind your dashboard. So, therefore, for dashboard presentations that really dig into what’s happening across the enterprise, the business should be more than willing to support SOA.
Rich Seeley recently posted some perspectives on Oracle’s Business Process Management 10g Release 3, a business intelligence dashboard that enables business users to customize their view of data coming from Oracle’s SOA-related technologies, including Complex Event Processing, ESB and BPEL.
The dashboard puts capabilities right in front of decision makers’ eyes. They may not entirely get SOA, but they clearly see the analysis of this quarter’s sales results and inventory levels displayed in front of them.
Roy Schulte, vice president of Gartner, is quoted as observing that the executive dashboard may be among the keys to bridging SOA and Business Process Management. “Business users may not understand SOA, BPM, CEP or XTP [eXtreme Transaction Processing], but they know what they want to see on their dashboard and they may be willing to fund back end architecture and development projects to get more information faster.”
Of course, if it’s nothing but bad news being displayed on the dashboard, the executive may just want to throw the whole thing out the window.
October 28th, 2008
Why ‘Event Driven Architecture’ is more than ‘Complex Event Processing’
Today’s mantra:
“EDA is an architecture; CEP is a technique…”
Jack van Hoof, who has written extensively on all things EDA, says there have been one too many instances in which the terms EDA (Event Driven Architecture) and CEP (Complex Event Processing) have been used almost interchangeably. He provides a clarification:
“CEP is a way of processing messages (fair enough to name these messages “events”)…. But EDA is about how business events drive the overall architecture of the IT-systems and it is about how these events should be modeled. EDA it is not primarily about the ability to process and correlate streams of thousands of messages per second…”
Just as SOA addresses the architectural aspect of service orientation for the business, EDA addresses the architectural aspect of how events should be processed for the business, Jack says. “CEP is just one among these aspects.” Please, he adds, EDA “EDA does not only deal with complex events (correlations) but also with simple events.”
“So CEP is not EDA, EDA is more than CEP. Promoting CEP as being EDA is far too simple. And yet that is what is happening in the current IT space.”
October 28th, 2008
Analyst: ‘SOA is working’
There’s been plenty of debate about how well SOA has been working for organizations.
Not every service should be expected to deliver business value
Count Forrester’s Randy Heffner is the pro-SOA camp. In an recent interview with ebizQ’s Peter Schooff, Heffner says most large organizations and at least half of small to medium businesses have adopted SOA, and most are satisfied with their progress so far. “For three years running, about 70% of current users have said they’ll do more SOA and only one percent or two percent will cut back. You don’t get a large consensus doing more of something unless it’s working,” he said.
Of course, not every service created under the umbrella of SOA delivers business value. In fact, some services are simply not meant to deliver value, Heffner said. Some services simply have more local or tactical purposes. “Many in the industry have what I like to call a ‘flat model of services.’ To them, a service is a service, is a service. Just put them all in the registry for people to search on. But it’s much better to realize that in terms a strategic business value, some services are better and more valuable than others.”
Heffner says there are actually three types of services:
Infrastructure services are technical utilities functions;
Applications services deal with connecting siloed applications; and
Business services, which embody organizations, major business transactions, and capabilities.
It’s the business services that will deliver the value, while the others enhance IT efficicency. “They’re things that business people care about as they design your business processes like submit orders, or convert accounts, or get customer history, or whatever it might be,” Heffner said. “A step in a business process, basically.”
Heffner also talked about the rise of REST-based services, or in broader terms, Web-Oriented Architecture (WOA). However, WOA is way too immature for enterprises at this point, he pointed out. “There are clearly scenarios where the lighter weight REST or WOA approaches are beneficial but the biggest limitation with them is that there’s no industry standard interoperability profiles for high quality of service with REST and WOA.”
October 27th, 2008
First 2009 SOA predictions arrive; see fading of the hype
Dave Linthicum has beat everybody to the punch and issued his predictions for the year ahead in service oriented architecture.
These are good predictions, my thoughts added:
1. The interest in cloud computing will drive many enterprises toward SOA.
Agreed. Cloud computing and software-as-a-service (SaaS) will make SOA real for many business users. In the same vein, user-driven, self-service functionality, such as mashups, will also put SOA more front and center to the business.
2. The explosion in PaaS (platform-as-a-service) will leave many enterprise architects and CIOs scratching their heads.
Agreed, with additional thoughts. PaaS-style offerings may take some of the technical headaches out of IT infrastructure, and make things a lot cheaper. Integration-as-a-Service will offer a compelling alternative for SOA efforts as well.
3. The economy will recover, but most enterprises out there will focus on cost reduction.
Agreed. But the focus on IT cost reduction has never let up since the 2001 downturn. This is where SOA can potentially pack a punch.
4. There will be a larger focus on inter-domain SOA technology, or highly scalable and secure middleware technology that will provide scalable servic