Technology -> Infrastructure
By: Dana Gardner, Principal Analyst, Interarbor Solutions
Published: 4th March 2013
Copyright Interarbor Solutions © 2013
We recently assembled a panel of enterprise architecture (EA) experts to explain how such simultaneous and complex trends as big data, cloud computing, security, and overall IT transformation can be helped by the combined strengths of The Open Group Architecture Framework (TOGAF®) and the ArchiMate® modeling language.
The panel consisted of Chris Forde, General Manager for Asia-Pacific and Vice President of Enterprise Architecture at The Open Group; Iver Band, Vice Chair of The Open Group ArchiMate Forum and Enterprise Architect at The Standard, a diversified financial services company; Mike Walker, Senior Enterprise Architecture Adviser and Strategist at HP and former Director of Enterprise Architecture at Dell; Henry Franken, the Chairman of The Open Group ArchiMate Forum and Managing Director at BIZZdesign, and Dave Hornford, Chairman of the Architecture Forum at The Open Group and Managing Partner at Conexiam. I served as the moderator.
This special BriefingsDirect thought leadership interview series comes to you in conjunction with The Open Group Conference recently held in Newport Beach, California. The conference focused on "big data—the transformation we need to embrace today." [Disclosure: The Open Group and HP are sponsors of BriefingsDirect podcasts.]
Here are some excerpts:
Gardner: Is there something about the role of the enterprise architect that is shifting?
Walker: There is less of a focus on the traditional things we come to think of EA such as standards, governance and policies, but rather into emerging areas such as the soft skills, business architecture, and strategy.
To this end I see a lot in the realm of working directly with the executive chain to understand the key value drivers for the company and rationalize where they want to go with their business. So we're moving into a business-transformation role in this practice.
At the same time, we've got to be mindful of the disruptive external technology forces coming in as well. EA can’t just divorce from the other aspects of architecture as well. So the role that enterprise architects play becomes more and more important and elevated in the organization.
Two examples of this disruptive technology that are being focused on at the conference are big data and cloud computing. Both are providing impacts to our businesses, not because of some new business idea but because technology is available to enhance or provide new capabilities to our business. The EA’s still do have to understand these new technology innovations and determine how they will apply to the business.
We need to get really good enterprise architects—it’s difficult to find good ones. There is a shortage right now, especially given that a lot of focus is being put on the EA department to really deliver sound architectures.
Gardner: We've been talking a lot here about big data, but usually that's not just a standalone topic. It's big data and cloud, cloud, mobile and security.
So with these overlapping and complex relationships among multiple trends, why is EA and things like the TOGAF framework and the ArchiMate modeling language especially useful?
Band: One of the things that has been clear for a while now is that people outside of IT don't necessarily have to go through the technology function to avail themselves of these technologies any more. Whether they ever had to is really a question as well.
One of things that EA is doing, and especially in the practice that I work in, is using approaches like the ArchiMate modeling language to effect clear communication between the business, IT, partners and other stakeholders. That's what I do in my daily work, overseeing our major systems modernization efforts. I work with major partners, some of which are offshore.
I'm increasingly called upon to make sure that we have clear processes for making decisions and clear ways of visualizing the different choices in front of us. We can't always unilaterally dictate the choice, but we can make the conversation clearer by using frameworks like the TOGAF standard and the ArchiMate modeling language, which I use virtually every day in my work.
Hornford: The fundamental benefit of these tools is the organization realizing its capability and strategy. I just came from a session where a fellow quoted a Harvard study, which said that around a third of executives thought their company was good at executing on its strategy. He highlighted that this means that two-thirds are not good at executing on their strategy.
If you're not good at executing on your strategy and you've got big data, mobile, consumerization of IT and cloud, where are you going? What's the correct approach? How does this fit into what you were trying to accomplish as an enterprise?
An enterprise architect that is doing their job is bringing together the strategy, goals and objectives of the organization. Also, its capabilities with the techniques that are available, whether it's offshoring, onshoring, cloud, or big data, so that the organization is able to move forward to where it needs to be, as opposed to where it's going to randomly walk to.
Forde: One of the things that has come out in several of the presentations is this kind of capability-based planning, a technique in EA to get their arms around this thing from a business-driver perspective. Just to polish what Dave said a little bit, it's connecting all of those things. We see enterprises talking about a capability-based view of things on that basis.
Gardner: Let's get a quick update. The TOGAF framework, where are we and what have been the highlights from this particular event?
Hornford: In the last year, we've published a minor upgrade for TOGAF version 9.1 which was based upon cleaning up consistency in the language in the TOGAF documentation. What we're working on right now is a significant new release, the next release of the TOGAF standard, which is dividing the TOGAF documentation to make it more consumable, more consistent and more useful for someone.
Today, the TOGAF standard has guidance on how to do something mixed into the framework of what you should be doing. We're peeling those apart. So with that peeled apart, we won't have guidance that is tied to classic application architecture in a world of cloud.
What we find when we have done work with the Banking Industry Architecture Network (BIAN) for banking architecture, Sherwood Applied Business Security Architecture (SABSA) for security architecture, and the TeleManagement Forum, is that the concepts in the TOGAF framework work across industries and across trends. We need to move the guidance into a place so that we can be far nimbler on how to tie cloud with my current strategy, how to tie consumerization of IT with on-shoring?
Franken: The ArchiMate modeling language turned two last year, and the ArchiMate 1.0 standard is the language to model out the core of your EA. The ArchiMate 2.0 standard added two specifics to it to make it better aligned also to the process of EA.
According to the TOGAF standard, this is being able to model out the motivation, why you're doing EA, stakeholders and the goals that drive us. The second extension to the ArchiMate standard is being able to model out its planning and migration.
So with the core EA and these two extensions, together with the TOGAF standard process working, you have a good basis on getting EA to work in your organization.
Gardner: Mike, fill us in on some of your thoughts about the role of information architecture vis-à-vis the larger business architect and enterprise architect roles.
Walker: Information architecture is an interesting topic in that it hasn’t been getting a whole lot of attention until recently.
Information architecture is an aspect of enterprise architecture that enables an information strategy or business solution through the definition of the company's business information assets, their sources, structure, classification and associations that will prescribe the required application architecture and technical capabilities.
Information architecture is the bridge between the business architecture world and the application and technology architecture activities.
The reason I say that is because information architecture is a business-driven discipline that details the information strategy of the company. As we know, and from what we’ve heard at the conference keynotes like in the case of NASA, big data, and security presentations, the preservation and classification of that information is vital to understanding what your architecture should be.
From an industry perspective, this is one of the least matured, as far as being incorporated into a formal discipline. The TOGAF standard actually has a phase dedicated to it in data architecture. Again, there are still lots of opportunities to grow and incorporate additional methods, models and tools by the enterprise information management discipline.
Enterprise information management not only captures traditional topic areas like master data management (MDM), metadata and unstructured types of information architecture but also focuses on the information governance, and the architecture patterns and styles implemented in MDM, big data, etc. There is a great deal of opportunity there.
From the role of information architects, I’m seeing more and more traction in the industry as a whole. I've dealt with an entire group that’s focused on information architecture and building up an enterprise information management practice, so that we can take our top line business strategies and understand what architectures we need to put there.
This is a critical enabler for global companies, because oftentimes they're restricted by regulation, typically handled at a government or regional area. This means we have to understand that we build our architecture. So it's not about the application, but rather the data that it processes, moves, or transforms.
Gardner: Up until not too long ago, the conventional thinking was that applications generate data. Then you treat the data in some way so that it can be used, perhaps by other applications, but that the data was secondary to the application.
But there's some shift in that thinking now more toward the idea that the data is the application and that new applications are designed to actually expand on the data’s value and deliver it out, to mobile tiers perhaps. Does that follow in your thinking that the data is actually more prominent as a resource perhaps on par with applications?
Walker: You're spot on, Dana. Before the commoditization of these technologies that resided on premises, we could get away with starting at the application layer and work our way back because we had access to the source code or hardware behind our firewalls. We could throw servers out, and we used to put the firewalls in front of the data to solve the problem with infrastructure. So we didn’t have to treat information as a first-class citizen. Times have changed, though.
Information access and processing is now democratized and it’s being pushed as the first point of presentment. A lot of times this is on a mobile device and even then it’s not the corporate’s mobile device, but your personal device. So how do you handle that data?
It's the same way with cloud, and I’ll give you a great example of this. I was working as an adviser for a company, and they were looking at their cloud strategy. They had made a big bet on one of the big infrastructures and cloud-service providers. They looked first at what the features and functions that that cloud provider could provide, and not necessarily the information requirements. There were two major issues that they ran into, and that was essentially a showstopper. They had to pull off that infrastructure.
The first one was that in that specific cloud provider’s terms of service around intellectual property (IP) ownership. Essentially, that company was forced to cut off their IP rights.
As you know, IP is a big business these days, and so that was a showstopper. It actually broke the core regulatory laws around being able to discover information.
So focusing on the applications to make sure it meets your functional needs is important. However, we should take a step back and look at the information first and make sure that for the people in your organization who can’t say no, their requirements are satisfied.
Gardner: Data architecture is it different from EA and business architecture, or is it a subset? What’s the relationship, Dave?
Hornford: Data architecture is part of an EA. I won’t use the word subset, because a subset starts to imply that it is a distinct thing that you can look at on its own. You cannot look at your business architecture without understanding your information architecture. When you think about big data, cool. We've got this pile of data in the corner. Where did it come from? Can we use it? Do we actually have legitimate rights, as Mike highlighted, to use this information? Are we allowed to mix it and who mixes it?
When we look at how our business is optimized, they normally optimize around work product, what the organization is delivering. That’s very easy. You can see who consumes your work product. With information, you often have no idea who consumes your information. So now we have provenance, we have source and, as we move for global companies, we have the trends around consumerization, cloud and simply tightening cycle time.
Gardner: Of course, the end game for a lot of the practitioners here is to create that feedback loop of a lifecycle approach, rapid information injection and rapid analysis that could be applied. So what are some of the ways that these disciplines and tools can help foster that complete lifecycle?
Band: The disciplines and tools can facilitate the right conversations among different stakeholders. One of the things that we're doing at The Standard is building cadres equally balanced between people in business and IT.
We're training them in information management, going through a particular curriculum, and having them study for an information management certification that introduces a lot of these different frameworks and standard concepts.
We want to create these cadres to be able to solve tough and persistent information management problems that affect all companies in financial services, because information is a shared asset. The purpose of the frameworks is to ensure proper stewardship of that asset across disciplines and across organizations within an enterprise.
Hornford: The core is from the two standards that we have, the ArchiMate standard and the TOGAF standard. The TOGAF standard has, from its early roots, focused on the components of EA and how to build a consistent method of understanding of what I'm trying to accomplish, understanding where I am, and where I need to be to reach my goal.
When we bring in the ArchiMate standard, I have a language, a descriptor, a visual descriptor that allows me to cross all of those domains in a consistent description, so that I can do that with traceability. When I pull in this lever or I have this regulatory impact, what does it hit me with, or if I have this constraint, what does it hit me with?
If I don’t do this, if I don’t use the framework of the TOGAF standard, or I don’t use the discipline of formal modeling in the ArchiMate standard, we're going to do it anecdotally. We're going to trip. We're going to fall. We're going to have a non-ending series of surprises, as Mike highlighted.
"Oh, terms of service. I am violating the regulations. Beautiful. Let’s take that to our executive and tell him right as we are about to go live that we have to stop, because we can't get where we want to go, because we didn't think about what it took to get there." And that’s the core of EA in the frameworks.
Walker: To build on what Dave has just talked about and going back to your first question Dana, the value statement on TOGAF from a business perspective. The businesses value of TOGAF is that they get a repeatable and a predictable process for building out our architectures that properly manage risks and reliably produces value.
The TOGAF framework provides a methodology to ask what problems you're trying to solve and where you are trying to go with your business opportunities or challenges. That leads to business architecture, which is really a rationalization in technical or architectural terms the distillation of the corporate strategy.
From there, what you want to understand is information—how does that translate, what information architecture do we need to put in place? You get into all sorts of things around risk management, etc., and then it goes on from there, until what we were talking about earlier about information architecture.
If the TOGAF standard is applied properly you can achieve the same result every time, That is what interests business stakeholders in my opinion. And the ArchiMate modeling language is great because, as we talked about, it provides very rich visualizations so that people cannot only show a picture, but tie information together. Different from other aspects of architecture, information architecture is less about the boxes and more about the lines.
Forde: Building on what Dave was saying earlier and also what Iver was saying is that while the process and the methodology and the tools are of interest, it’s the discipline and the quality of the individuals doing the work.
Iver talked about how the conversation is shifting and the practice is improving to build communications groups that have a discipline to operate around. What I am hearing is implied, but actually I know what specifically occurs, is that we end up with assets that are well described and reusable.
And there is a point at which you reach a critical mass that these assets become an accelerator for decision making. So the ability of the enterprise and the decision makers in the enterprise at the right level to respond is improved, because they have a well disciplined foundation beneath them.
A set of assets that are reasonably well-known at the right level of granularity for them to absorb the information and the conversation is being structured so that the technical people and the business people are in the right room together to talk about the problems.
This is actually a fairly sophisticated set of operations that I am discussing and doesn't happen overnight, but is definitely one of the things that we see occurring with our members in certain cases.
Hornford: I want to build on that what Chris said. It’s actually the word "asset." While he was talking, I was thinking about how people have talked about information as an asset. Most of us don’t know what information we have, how it’s collected, where it is, but we know we have got a valuable asset.
I'll use an analogy. I have a factory some place in the world that makes stuff. Is that an asset? If I know that my factory is able to produce a particular set of goods and it’s hooked into my supply chain here, I've got an asset. Before that, I just owned a thing.
I was very encouraged listening to what Iver talked about. We're building cadres. We're building out this approach and I have seen this. I'm not using that word, but now I'm stealing that word. It's how people build effective teams, which is not to take a couple of specialists and put them in an ivory tower, but it’s to provide the method and the discipline of how we converse about it, so that we can have a consistent conversation.
When I tie it with some of the tools from the Architecture Forum and the ArchiMate Forum, I'm able to consistently describe it, so that I now have an asset I can identify, consume and produce value from.
Forde: And this is very different from data modeling. We are not talking about entity relationship, junk at the technical detail, or third normal form and that kind of stuff. We're talking about a conversation that’s occurring around the business context of what needs to go on supported by the right level of technical detail when you need to go there in order to clarify.
We automatically stop accepting comments 180 days after a post is published. If you would like to know more about this subject, please contact us and we'll try to help.
Published by: electronicdawn Ltd.