Hype over Cloud Computing and SOA have inverted

It is hard to measure hype. We can detect it by well trained B.S. sensors, but how we go about quantifying it is another question.

Take for example the hype around SOA as a solution for all enterprise needs. IT professionals have been reading about SOA for over a decade, and with the rise of web services SOA-type solutions became very prevalent and were quickly making solid contributions to enterprise missions. But at some point the term became over hyped, with marketing teams from just about every IT provider touting their SOA solution and even some IT professionals starting to throw the term around with little regard for its meaning. But those general observations on hype are hard to measure. We can say, however, that for designers, the term SOA has real meaning but the hype can be distracting.

Many of us have been watching the same phenomenon with the term “Cloud Computing.” This construct has had real meaning to enterprise technologists, architects, planners and other IT professionals for years. But the hype and resulting misuse of the term is getting distracting.

In July 2009 we wrote of a coming “cross-over in hype” between cloud computing and SOA. It seemed hype in the marketplace over SOA was declining (and that’s good!) but hype over Cloud Computing was increasing.  As one indicator we presented a chart from Google Trends that showed that searches on the term SOA were tending down as searches on the term CloudComputing were increasing.

An update of the chart using the same search terms points to a crossover point occurring somewhere in the last quarter of 2010.

These charts don’t really measure hype, which remains a very hard to quantify construct, but it definitely shows the cloud computing concept seeming to emerge from no where and the interest, as reflected in searches, increasing over SOA. So, it doesn’t take much of a leap to observe that the hype around these two concepts is shifting.  If SOA is no longer the hot new thing and if cloud is the hot new thing then the marketing engines are going to reflect that.

Does that fit with your experience?

The bottom line: Mission should drive designs and architects should base designs off of an understanding of those missons informed by the realities of legacy systems and emerging technologies. Beware the hype on all topics, especially cloud computing. Cloud concepts are critically important to designs, but when something has so much hype around it designers should ensure solid definitions of the term are in place so all understand what is being talked about.

The FBI Reading Room is a great source of information

I was recently treated to a demonstration of a mission-focused IT solution that made me proud of what technologists at FBI can do. What I saw was a best in class integration of search, discovery and analysis tools known as DIVS (Data Integration and Visualization System). DIVS includes a framework that allows multiple tools, COTS and GOTS, to interoperate in ways incredibly supportive of the most significant missions in the national security space.

For those that have had the pleasure of meeting the leaders of the IT team there at FBI, or for any who interact with the technological wizards of the FBI’s Science and Technology Branch, you are probably not surprised that they are fielding best in class solutions like these. The FBI attracts top class talent that knows how to field working enterprise solutions. I believe we will be reading about many more solutions like these coming out of the FBI in the near term.

Related articles

• FBI Recasts Sentinel As Model Of Agility (informationweek.com)

The 2010 DoDIIS Worldwide Conference will be held 23-27 May in Phoenix Arizona.  This venue is hosted by the Defense Intelligence Agency’s Directorate for Information Management and Chief Information Officer.  The focus is on topics relevant to mission success for the Department of Defense Intelligence Information Systems (DoDIIS).  DoDIIS is all about facilitating the collection, analysis and presentation of intelligence to tactical users and policy makers in a secure and timely manner.

This year’s agenda is particularly impressive.  The agenda (which is pasted below as a reference) includes world class industry speakers like Joseph Tucci, CEO of EMC, and Dr. Steven Armentrout, CEO of Parabon.  Government speakers include the leaders of DoD intelligence IT like Grant Schneider, Roy Apseloff, and Casey Henson.  The national intelligence IT team will also be present in force, including a panel chaired by ODNI CIO Priscilla Guthrie.

The event will also hold an expo floor with 100′s of advanced technologies and mission-focused solutions being demonstrated.  It is my intent to spend time with 100% of the companies that have decided to participate in the expo, this is a great way to take the pulse of technology that is ready for insertion into large enterprises and a great way to generate topics for future blog posts and for evaluation at CTOlabs.com

Agenda:

24 May 2010

25 May 2010:

26 May 2010:

27 May 2010:

One of my friends and mentors, Bill Vass, has consistently advised IT professionals to understand, respect and use the power of well formed IT architecture.  Bill has often reminded me and others that architecture is design.  If you have poor design, or if you have good design that is ignored, then the architecture is worthless or even counter productive.  If you have good, actionable design that is focused on mission needs and used by enterprise decision-makers then IT is better able to deliver workable solutions.  Bill also emphasizes that well done architecture begins with the end in mind.

That last point means, that before building your design you need to know what your mission requirements are.  And you need a good relationship with your customers and how they serve the mission.

If there was ever any doubt of that, I’d like to point out a great example highlighted by Monty Python in a skit named… “Architect Sketch.” In it John Cleese plays an architect who seems to have created his design without an approved business architecture. The result is a design that does not meet the intended mission needs. (In the same sketch, Eric Idle, seems to have a design that reflects mission needs, but shows signs that it might not be executable).

The sketch is a great way to drive home many lessons in IT architecture, don’t you think?  Here are a few that come to mind:

• IT customers should very clearly spell out the requirements that need solving (by the way, most are are not very good at this)
• IT architects must ensure they are designing to mission needs
• Sometimes poor technical designs are picked because they are the only thing that comes close to meeting the mission
• Sometimes designs are picked because of relationships

If you are a customer of IT is is your duty to describe requirements in clear, understandable ways so the IT architects have a chance of designing the right solution.

If you are an IT architect you have a duty to ensure you are designing things to meet mission needs.  You don’t design something just because that is what you are good at (John Cleese’s character said “You see I mainly design slaughter houses.”).   And if you design something that customers reject in design review, don’t do a Cleese on them.  Although he does have a way with words when it comes to pushing back on user concerns:

“Well, of course, this is just the sort of blinkered philistine pig-ignorance I’ve come to expect from you non-creative garbage. You sit there on your loathsome spotty behinds squeezing blackheads, not caring a tinker’s cuss for the struggling artist. You excrement, you whining hypocritical toadies with your colour TV sets and your Tony Jacklin golf clubs and your bleeding masonic secret handshakes.”

(That kind of in-your-face interaction with customers is probably not conducive to a productive, healthy, long-term relationship with your customers).

Which raises another point this skit humorously highlights.  In the end, the customer choose the poor technical design over the good one because it was closer to meeting the mission, and because the architect had a relationship with the customer, as evidenced by the well executed masonic secret handshake.

Lesson overall: Good enterprise IT is always about people and the mission.  When enterprise IT is designed without a good understanding of the mission, disaster can result.

Most enterprise technologists should see a continued payoff of the hard work in planning, architecture, documentation, development and configuration work that has been occurring over the last several years.  Enterprise technologists were building Service Oriented Architectures (SOA) long before SOA was over-hyped.  And most enterprise technologists I know were investigating constructs of scalable, elastic Cloud Computing capabilities long before that became the dominate theme in trade journals, conferences and tech blogging.  With all the hard work and progress seen in enterprises to date we could be in store for some very positive improvements in 2010.

There are other mitigating factors to watch, however.  Some that come immediately to mind are the steady pressures to reduce budgets and the constantly increasing security challenges.  Mission requirements are also continuing to accelerate.

So what will the net assessment be by the end of 2010? There are many variables at play here, but in theory the greatest determinant of technological outcomes are the leaders of the IT enterprise.  What you decide will have a great impact on how well your mission is served.

As an aid to your decision making, the CTOvision.com site tracks the megatrends we see sweeping across the IT landscape.  The following is an update on six we believe to be of significant relevance to the enterprise Chief Technology Officer (CTO) and Chief Information Officer (CIO).  They are:

• Convergence and trends towards unified communications and user empowerment
• The continued dominance of American IT
• Increasing open development of software and hardware
• Cloud Computing and massive ingest/parsing of data
• Green IT and support to total mission effectiveness
• Increasing pace of technology improvement/development

Convergence and trends towards unified communications and user empowerment. Consumerization and its impact on IT development has been a trend for the last five years and it shows no sign of slowing. All IT around the globe is being impacted by this trend, and IT development is focusing increasingly on consumers vice government or enterprises.  The good news here is that enterprise employees are increasingly becoming technology savvy power users that should be fast studies when it comes to learning the power of enterprise IT. The bad news is many enterprise users will have very high, frequently unrealistic expectations of what their enterprise IT will be able to provide for them, since enterprises have special security and regulatory (as well as budgetary) concerns that many home uses do not have.  Another important impact of this trend is that most enterprise users are becoming very used to social media, which can enable enterprises to support mass collaboration on problems.

Consumers are also a big reason the IT industry is fielding increasing location-aware capabilities, and many of these will be of use in enterprise solutions (one we are watching closely is SenseNetworks).

Individual enterprise users are also the reason IT departments must focus on information discovery capabilities vice the old style information search tools. Users are demanding this and other productivity advancements.

The continued dominance of American IT. The fact that US stockholders own and US citizens run most all enterprise IT firms is something we should all be proud of.  Our American way of education and competition has generated great firms that dominate, for now, the IT landscape (see the CTOvision.com Tech Titan List). For enterprises with special national security concerns this can be one factor that helps in assessments of threats to IT supply chain attacks.  But also of importance, the leadership of these firms and their corporate headquarters are all very open to interactions from enterprise IT leaders from the federal government, and in every interaction I’ve ever had with them they are all very appreciative of the missions supported by federal IT leaders.  The availability and proximity of these great firms is something our federal enterprise should take continued advantage of.

Increasing open development of software and hardware. All major IT firms, including the powerhouses that produce proprietary software and hardware, are now embracing the open source movement.  Even Microsoft has an open source strategy that applies to some of their offerings.  The open source community has long benefited from the developer talent in big companies that help produce and further code in Linux (see Red Hat for a supported variety), OpenOffice and many other open source solutions.  We can all expect proprietary software will be the most full featured software (for many market reasons), but even that will be built in a way that works well with open source. The benefit to enterprises is a wider range of choices in solutions, and in many cases an ability to field solutions faster and with more security and lower cost.  We can also expect all federal enterprises will find ways to enhance internal collaboration on software development projects, for example, the Forge.mil collaborative development capability.

Cloud Computing and massive ingest/parsing of data. After years of planning and hard work and design enterprise technologists have finally begun to see benefit from the concept of “private clouds,” where the scalable, elastic access to resources leveraged by public clouds can now be used in more secure enterprise infrastructures.  And the cloud computing trends in the IT sector have resulted in a great suite of interoperable capabilities ready to improve private cloud security, reliability and performance.  Private clouds also strongly support continued requirements for large scale ingest and parsing of data to support enterprise missions, and they can do this in a way that makes disaster preparedness and recovery much easier to plan for.

Green IT and support to total mission effectiveness. The movement to Green IT first began as a way to do the right thing, with government and industry realizing the PC’s and monitors being fielded should be built in more energy efficient ways. Perhaps the most visible activity in this area is Energy Star, a government-backed program designed to help businesses and individuals save energy.  But many other activities have arose including an executive order requiring all federal IT efforts to support strict new Green IT rules.  The cost of energy has also been a continual driver.  As enterprises look for ways to save power they are also looking at travel budgets and the cost of sending people to distant meetings and this is placing more demand on capabilities that enable collaboration from a distance.  Energy efficiency is also another driver of telecommuting, which will put pressure on IT departments to enable secure access to enterprise clouds from anyplace a worker is.

Increasing pace of technology improvement/development. The CTO’s of federal enterprise are a busy bunch.  Not only do they have to articulate standards and guidance for their enterprise, but they must continually track the many changes coming out of IT companies big and small.  There is never enough time to do this, and all indications are that the speed of technology advancement will only increase.  All enterprise technologists should consider the mechanisms they can put in place to learn of and evaluate technologies which may be of interest.  My hope is that one of your resources in doing that is my blog at http://ctovision.com This site tracks megatrends, including the six in this post. We also list firms big and small that are fielding capabilities which can dramatically enhance mission effectiveness, and do so in order to provide advanced warning for enterprise technologists who must maintain awareness of what is coming.

These trends are relevant to all, and an awareness of them should help your planning process.  They call for an acceleration of green IT, virtualization systems, social media, collaboration tools and new user-focused devices into the enterprise.  They also call for more direct dialog with users since many of these capabilities can transform their way of work.  And they call for enterprise technologists from every agency to continue to scan the horizon to look for what’s new.  That need will likely never stop.

Adobe Flex 3 Bible

Adobe Flex is a software development approach and environment which enables the production of cross-platform Rich Internet Applications.  Adobe Integrated Runtime (AIR) is the cross-platform runtime environment used by Flex (as well as Flash, HTML, JavaScript and AJAX) for deployment of desktop applications.

I’ve long known the power of Flex and AIR, and have long respected the smart strategy of enabling developers and enterprises to decide when a solution should be one that will run in the browser or on the client. But to really learn the power of Flex and AIR I decided to try coding a few examples.  After ordering and working my way through a few books (including David Gassner’s  Flex 3 Bible) and building a few sample programs, I decided to create an application designed to serve friends in the national security community. I thought doing so would force me to learn more details of the Flex AIR approaches and could also result in me producing an application of use to some military planners.

Some lessons:

• Flex is easy to learn. If you know Visual Basic you will enjoy and appreciate the graphic design capabilities of Flex.   If you know object oriented languages you will feel right a home with Flex’s use of functions, objects and events.
• Flex is easy to test and debug.  There are great visuals to help you spot mistakes. And it is easy to repeatedly run your work in progress.
• Flex is easy to deploy, in both desktop and web editions.  Use of Flex and AIR makes preparation, packaging and deployment really fast.
• Overall, creating an application is fast.  I went from idea to application in a matter of days.  Part of that is because this is a simple application, but part of it is because Flex is so easy and fast to work with.
• I believe every enterprise should have AIR on every client.

You can download Flex for free from Adobe.  The download comes with a 60 day license.  Developers in between work can apply for an extended free license.  To download yours see: http://www.adobe.com/products/flex/

In a coming post I’ll provide links to the first version of my application.  I’ve sent it to a couple friends to let them beat it up before making it totally public.  And I am thinking of adding a few other features to make it a bit more polished before posting it.  But standby, it is coming soon.

Till then, please ensure your computer is running the most current version of AIR.  Do that by downloading:

http://get.adobe.com/air/

Air runs on Windows, Mac and Linux OS’s, making it really cross-platform.

Cloud Computing is one of the many things enterprise CIOs, CTOs and other engineers will master in delivering capability.  I believe in the power of new Cloud Computing technologies and concepts and think we should all continue our focus there.

I have said, and still say, the same thing about design approaches like Service Oriented Architecture (SOA).  The constructs, methods and models of SOA are good practices that result in good designs for enterprises.   It is smart to separate data from application logic and smart to enable agility and mashups the way good SOA design does.

But a key problem with SOA was the hype associated with it.  Everyone started using the term the way they wanted to.  And every IT vendor came out with their definition of SOA.  And every system integrator tried to sell us buckets of their SOA.  And every trade journal ran SOA articles.  And too many CTOs, myself included, said too much to our customer/users about the promise of SOA and that contributed to our users increasing their expectations over what SOA would deliver.   The result of all of that, the over hype on SOA sometimes caused distractions.  The cure: CTOs focused on what SOA meant for their specific organizations.  It is still a great construct and by focus it is delivering value.

There are now clear indications that the hype around Cloud Computing is going to ramp way up.  In fact, a quick search on Google Trends indicates searches on the topic of Cloud Computing, while still lower than searches on the topic of SOA, are rapidly gaining while searches on SOA are declining.  The two are headed for a cross-over.

The graph below reflects this (a live version of this graph is at Google Trends).

Here are my views of the “So What” of this information.  For the Chief Technology Officer or Chief Information Officer or other enterprise technologist who must really deliver, I’d recommend you brace yourself for a flood of hype.  You have probably already seen indications of this.  But if trends continue the amount of views you (and your bosses) will be asked to consider and deal with will accelerate dramatically.  Your defense against this will be the strength of your own position in Cloud Computing.  Therefore, I strongly recommend you personally think through what Cloud Computing means to your enterprise.  Also, think through your definition of cloud computing.  Hopefully the NIST definition will suit your use, since the more people who form up on that the better.

And you may want to think through which Cloud Computing related events you will attend and speak at.

And you may want to think through you who personally believe the strongest IT players are in the Cloud Computing world.  I have my list.  Let me know if you want to compare notes sometime.

Any thoughts?

“The nice thing about standards is that you have so many to choose from” teaches Andrew S. Tanebaum in his classic text on Computer Networks.

His point is so very important for CTOs to noodle over.  Every vendor, integrator and PM will tell you they are following standards.  GREAT!   But so what?  The point is they need to be following the standards you expect them to follow and they need to be building to your vision by following your guidance.

For CTOs, standards are things that everyone tries to deviate from.  They are created to make life easier and when appropriately developed and used/implemented they can significantly enhance the overall
functionality and performance of all elements of the technology stack.  Even companies which develop proprietary, closed source code like standards, but generally those companies tend to ignore standards
whenever it is convenient since doing so can help them lock others out of key market areas.

As an aid to thinking through standards, I’d recommend CTO’s turn to the standard for online encyclopedic information, Wikipedia.  Wikipedia has a couple great entries on standards.  The primary entry is at:  http://en.wikipedia.org/wiki/Standard.   From there we read that a standard is “It is usually a formal document that establishes uniform engineering or technical criteria, methods, processes and practices.”

Elsewhere on Wikipedia there is a list of all standards bodies.  Not every standards body deals with topics of interest to IT focused CTOs.  The list below is an extract from the Wikipedia article that only has those groups I think are working topics of interest to CTOs.

• 3GPP – 3rd Generation Partnership Project – Website
• 3GPP2 – 3rd Generation Partnership Project 2 – Website
• AIIM – Association for Information and Image Management – Website
• ANSI – American National Standards Institute
• DMTF – Distributed Management Task Force. develops and maintains standards for systems management of IT environments in enterprises and the Internet.
• Ecma International – Ecma International (previously called ECMA).  Computer standards with a business-like approach.
• GS1 – Global supply chain standards (identification numbers, barcodes, electronic commerce transactions, RFID) – Website
• IBTA – Infiniband Trade Association
• IEEE – Institute of Electrical and Electronics Engineers – Website
  
• IETF – Internet Engineering Task Force – Website
• ISO – International Organization for Standardization – Website
• ITU – The International Telecommunication Union – Website
  • ITU-R – ITU Radiocommunications Sector (formerly known as CCIR)
  • ITU-T – ITU Telecommunications Sector (formerly known as CCITT)
  • ITU-D – ITU Telecom Development (formerly known as BDT)
• Liberty Alliance – Liberty Alliance – Website
  
• Media Grid – Media Grid Standards Organization – Website
• OASIS – Organization for the Advancement of Structured Information Standards – Website
• OGC – Open Geospatial Consortium – Website
• OMA – Open Mobile Alliance – Website
• OGF – Open Grid Forum (merger of Global Grid Forum (GGF) and Enterprise Grid Alliance (EGA)) – Website
• TM Forum – Telemanagement Forum – TMF Website
• W3C – World Wide Web Consortium – Website
• WSA – Website Standards Association Website

I’d recommend a review of the list and a familiarization with the sites of the organizations above.  Staying fluent in the content on the sites above can help you provide valuable context and direction to development team and integrators, and can help in your dialog with IT vendors.

The term SOA (Service Oriented Architecture) has been around for quite a while, and it seems like everyone is using it.   But is everyone using it correctly?   I guess that is open to debate, but since everyone seems to be using it differently the odds are that enterprise CTOs will have to think through their own use of the term and may need to issue guidance to vendors, integrators and internal staff on how the term is to be used in the enterprise.  .

I wondered how many definitions of SOA there are out there today.  So I did a Google Search.  I wanted to limit my search to only those sites that claim to provide definitions.  So I launched a search for “SOA intitle:dictionary ”

Of course that search will not give precise results, I thought it would narrow my results down.

Thank goodness I narrowed my results down.  I only got 12,300 hits back from that search.

Are there really 12,300 different definitions of SOA?  It sure seems like it, and after scrolling through page after page of results from Google I can say that there is not enough precision in the existing definitions.

So, what is a CTO to do?  Here are some options:
1) you can accept a definition provided by your favorite integrator,
2) you can accept a definition provided by your favorite vendor,
3) you can pick one of the 12,300 dictionary sites on Google that have an answer, or,
4) you can make up your own.

In my last organization I needed to work with many integrators and many vendors, all of which had their own definition of SOA, and I had to work with many internal government mission partners, all of which used SOA the way they wanted to.   So I decided the only real option was to write my own definition.   I put the following in my guidance to the enterprise:

“SOA is a design for linking computational resources (principally applications and data) on demand to achieve the desired results for service consumers (which can be end users or other services).”

I was a little worried that there might be something wrong with that definition.  Maybe I left something out?  Or maybe I said something that would do unintentional damage to the enterprise or some of our projects.  So I decided to put that out into the public domain for others to shoot at.   I looked up wikipedia and searched for the term SOA.

At the time there was an entry for SOA, but it was small and did not actually have a definition.  So I created a section called “SOA Definition” at the SOA stub on Wikipedia and added my definition

Within hours someone had edited the entry to add the OASIS (Organization for the Advancement of Structured Information Standards) definition of SOA, which is:

“A paradigm for organizing and utilizing distributed capabilities
that may be under the control of different ownership domains. It
provides a uniform means to offer, discover, interact with and use
capabilities to produce desired effects consistent with measurable
preconditions and expectations.”

I like that definition too, but in my view it is a more technical description that is not always going to be good for the more strategic decision-makers in an enterprise.  I’ve decided to stick with my definition for everyday use, but to fall back to the OASIS one if I need to have a technical discussion.

The following is a review of some new capabilities at Google (these were all announced and discussed on the official Google blog):

Better and faster route planning on maps.  Google maps now has dragable directions, and you can also enter guidance and info on Google maps like “avoid highways” or calculate “time-in-traffic”. That’s pretty cool

A new operating system and open architecture for mobile phones has been announced. This is a comprehensive platform for mobile devices that should give a significant boost to innovation in the mobile market. A software development kit will be available 12 November.

A new architecture and framework for social network data exchange has been developed. This is known as “OpenSocial”. This is a set of common API’s that make it easy to create social software. The widespread adaptation of these API’s will make the net a more social place and will make it easier on humans like us. Global members of the OpenSocial community include MySpace, Engage.com, Friendster, hi5, Hyves, imeem, LinkedIn, Ning, Oracle, orkut, Plaxo, Salesforce.com, Six Apart, Tianji, Viadeo, and XING.

These are just three of a long line of capabilities coming out from these guys.  Capabilities are also continuously being added to other elements of the Google platform.  For example, Google documents is frequently being upgraded, and since the capabilities of Google Docs are delivered as a web service, the upgrade is automatic and painless.

Google has come a long long way from just a search company.

But there are more significant points for CTOs.  For example, since these many capabilties are just some recent examples, are we starting to become numb to the news of new things from Google?  Each of the items mentioned above as recent Google capabilities have the potential of being very disruptive to the current web services world.  But how can we analyze or assess the potential disruption if we have become so numb to the news?

And another thought:  What if the rate of innovation or even the rate of perceived innovation slows down from Google?  We have become hooked on the frequent announcement of incredible new capabilities.  What will we do when the announcements slow?

But, as an enterprise technologist, my biggest question is always “when will the capabilities being fielded by Google be available inside the enterprise?  On behalf of users everywhere, I hope the answer is “soon.”

Some references to the above:

http://googleblog.blogspot.com/2007/11/road-to-better-path-finding.html

http://google-latlong.blogspot.com/2007/06/its-click-drag-situation.html

http://www.openhandsetalliance.com/developers.html

http://googleblog.blogspot.com/2007/11/opensocial-makes-web-better.html

]]> http://ctovision.com/2007/11/google-upgrades-and-enhancements/feed/ 2