Listen to my latest podcast For EA in the Government, Just Follow the Money.
Feel free to send comments. I look forward to your feedback and suggestions for future podcasts.
Friday, April 23, 2010
Thursday, April 22, 2010
Meaningful SOA Metrics
Posted by
Tarak, TekNirvana.com
This is essentially the sequel of the previous discussion about service reuse on the ebizQ SOA forum. If we accept the premise that service reuse is not an optimal metric then what is?
Well, SOA is often touted as the "silver bullet" that will increase resusability, enhance maintainability, reduce rework, and so on. In that spirit, I think a metric of prime importance should be the Total Cost of Ownership (TCO) per Unit of Software Complexity (or Size) such as Function Points, Use Case Points, SLOC, etc. Assuming that you are already well down the path of SOA, a necessary first step would be to establish a historical baseline (pre-SOA) to compare against. The hope is to see current (SOA-based) measurements below the baseline with a continuous downward sloping trend.
* Originally posted on the ebizQ SOA forum on April 22, 2010.
Well, SOA is often touted as the "silver bullet" that will increase resusability, enhance maintainability, reduce rework, and so on. In that spirit, I think a metric of prime importance should be the Total Cost of Ownership (TCO) per Unit of Software Complexity (or Size) such as Function Points, Use Case Points, SLOC, etc. Assuming that you are already well down the path of SOA, a necessary first step would be to establish a historical baseline (pre-SOA) to compare against. The hope is to see current (SOA-based) measurements below the baseline with a continuous downward sloping trend.
* Originally posted on the ebizQ SOA forum on April 22, 2010.
Wednesday, April 21, 2010
Stevie "Wonders" about Cloud Computing
Posted by
Tarak, TekNirvana.com
I never thought I'd meet music legend Stevie Wonder. So imagine my surprise when he came to G&B yesterday and sat in the office right next to me almost all day! I finally gathered enough courage to take the few steps to introduce myself to him.
Amazingly, within the first few minutes our conversation led to Cloud Computing. He had heard about it but did not know what it really meant. I explained it to him comparing the evolution to the rise of electic utility companies. He got it and I know because he started finishing my sentences! We talked for quite a while about other things too: religion, politics, terrorism, and my (East) Indian heritage. He even sang a line from one of his songs!
A truly amazing experience indeed... After all, how many technologists can claim that they explained Cloud Computing to Stevie Wonder!
Thursday, April 15, 2010
Service Reuse... What are reasonable expectations?
Posted by
Tarak, TekNirvana.com
Today's ebizQ discussion on the SOA forum is around service reuse. To me, this is deja vu all over again. I remember this same discussion about libraries, procedures, objects, and components. Now it's about services.
In my opinion, service reuse is the wrong metric to be tracking with SOA. In addition to be being very subjective, it's a very tactical metric akin to saying that an "enterprise is secure because our firewall has no known vulnerabilities." While an amusing statistic, service reuse is not why an SOA should be implemented, nor is it the proper way to justify it. Then what makes the business case for SOA? That would be a tangible, demonstrable alignment to organizational and business objectives.
* Originally posted on the ebizQ SOA forum on April 15, 2010.
In my opinion, service reuse is the wrong metric to be tracking with SOA. In addition to be being very subjective, it's a very tactical metric akin to saying that an "enterprise is secure because our firewall has no known vulnerabilities." While an amusing statistic, service reuse is not why an SOA should be implemented, nor is it the proper way to justify it. Then what makes the business case for SOA? That would be a tangible, demonstrable alignment to organizational and business objectives.
* Originally posted on the ebizQ SOA forum on April 15, 2010.
Tuesday, April 13, 2010
Is Web 3.0 a Convergence of Cloud and Web?
Posted by
Tarak, TekNirvana.com
We've been having some interesting conversations on ebizQ recently. Today's topic was based on a recent article on ebizQ that pondered whether Web 3.0 was really just a convergence of the Web and the Cloud.
According to Tim Berners-Lee (paraphrased of course) Web 3.0 is an overlay of what he has been calling the "Semantic Web" on top of the existing Web 2.0. Nova Spivack defines Web 3.0 as a connective intelligence; connecting data, concepts, applications and ultimately people with the Semantic Web being a part of it but not the end all. Almost every other definition of Web 3.0 also makes some reference to a core component of "intelligence" as well.
The question about whether the convergence of Cloud and Web is Web 3.0 is interesting. After some thought, however, I believe that while Clouds will be an essential foundation to support 3.0, they are not the critical component of 3.0. Their role will be similar to others such as BI, SOA, AI, etc. but just as Web 2.0 is not Web + SOA, Web 3.0 will not be Web + Clouds. In fact, one could argue that clouds have always been around. After all, haven't we all depicted the Web as a "Cloud" at one time or another?
* Originally posted on the ebizQ Cloud Computing Forum on April 13, 2010.
According to Tim Berners-Lee (paraphrased of course) Web 3.0 is an overlay of what he has been calling the "Semantic Web" on top of the existing Web 2.0. Nova Spivack defines Web 3.0 as a connective intelligence; connecting data, concepts, applications and ultimately people with the Semantic Web being a part of it but not the end all. Almost every other definition of Web 3.0 also makes some reference to a core component of "intelligence" as well.
The question about whether the convergence of Cloud and Web is Web 3.0 is interesting. After some thought, however, I believe that while Clouds will be an essential foundation to support 3.0, they are not the critical component of 3.0. Their role will be similar to others such as BI, SOA, AI, etc. but just as Web 2.0 is not Web + SOA, Web 3.0 will not be Web + Clouds. In fact, one could argue that clouds have always been around. After all, haven't we all depicted the Web as a "Cloud" at one time or another?
* Originally posted on the ebizQ Cloud Computing Forum on April 13, 2010.
Thursday, April 8, 2010
Is SOA the Foundation of a Private Cloud?
Posted by
Tarak, TekNirvana.com
An interesting question posed by ebizQ's Joe McKendrick today on the SOA forum. His premise is that the purpose of SOA is to enable delivery and reuse of services across the enterprise, which are made available to end-user departments via a directory or registry/repository meshes with the concept of private cloud. Therefore, if you are building an SOA-based infrastructure, are you not also building a private cloud?
I think the question has been artificially constricted in that it should not be limited to the "private" cloud but ALL clouds in general.
In terms of the relationship between SOA and Clouds, I see them as complementary, convergent, and as essential constituents of a larger ecosystem. This is exactly what I talked about in my article titled The Cloud SOA Ecosystem on ebizQ back in October of 2009.
A few excerpts follow:
"The convergence of the cloud and SOA was bound to happen. Several independent events have led up to this perfect union. First, the emergence of the cloud as a viable business and innovation platform that is evident from the sheer number of vendors proclaiming "cloud" products and services and the wide scale adoption of the cloud by many Fortune 500 companies and even the U.S. Federal government. Second, the economic downturn has brought cost cutting initiatives, frugality, and operational efficiency back in fashion. Third, SOA, although undoubtedly a superior architectural style, has struggled to remain mainstream in the eyes of the business beyond the notion of "improving reusability" -- so much so that many technology pundits were even questioning whether SOA was dead!...
...As if the above events were not enough to bring them together, SOA and the cloud have a shared objective: improving the alignment of IT and business with the goal of justifying current and future IT investments, improving IT responsiveness to business change, and increasing operational efficiency. Both the cloud and SOA move the IT function beyond managing applications and infrastructure into a partnership with the business to provide agile solutions to solve dynamic business problems.
The convergence of the cloud and SOA goes beyond just simple alignment of common needs or shared objectives. It is a true symbiotic relationship where SOA fulfills the basic need of the cloud of an application architecture that aligns with and can support its own virtualized, multi-tenant, elastic being.”
Now, how about adding BPM to the mix?
* Originally posted on the ebizQ SOA forum on April 8, 2010
I think the question has been artificially constricted in that it should not be limited to the "private" cloud but ALL clouds in general.
In terms of the relationship between SOA and Clouds, I see them as complementary, convergent, and as essential constituents of a larger ecosystem. This is exactly what I talked about in my article titled The Cloud SOA Ecosystem on ebizQ back in October of 2009.
A few excerpts follow:
"The convergence of the cloud and SOA was bound to happen. Several independent events have led up to this perfect union. First, the emergence of the cloud as a viable business and innovation platform that is evident from the sheer number of vendors proclaiming "cloud" products and services and the wide scale adoption of the cloud by many Fortune 500 companies and even the U.S. Federal government. Second, the economic downturn has brought cost cutting initiatives, frugality, and operational efficiency back in fashion. Third, SOA, although undoubtedly a superior architectural style, has struggled to remain mainstream in the eyes of the business beyond the notion of "improving reusability" -- so much so that many technology pundits were even questioning whether SOA was dead!...
...As if the above events were not enough to bring them together, SOA and the cloud have a shared objective: improving the alignment of IT and business with the goal of justifying current and future IT investments, improving IT responsiveness to business change, and increasing operational efficiency. Both the cloud and SOA move the IT function beyond managing applications and infrastructure into a partnership with the business to provide agile solutions to solve dynamic business problems.
The convergence of the cloud and SOA goes beyond just simple alignment of common needs or shared objectives. It is a true symbiotic relationship where SOA fulfills the basic need of the cloud of an application architecture that aligns with and can support its own virtualized, multi-tenant, elastic being.”
Now, how about adding BPM to the mix?
* Originally posted on the ebizQ SOA forum on April 8, 2010
Tuesday, April 6, 2010
Should a Private Cloud Still Be Considered a Cloud?
Posted by
Tarak, TekNirvana.com
This deceivingly innocuous question sparked another lively debate (and some name calling) on the ebizQ Cloud Forum today. So, is a private cloud still a cloud?
Absolutely.
Let's not get hung up on names. The end game is to realize the benefits of the cloud business model. Yes, some options might be better at realizing a subset of benefits (such as cost reduction) than others but not all options are suitable for all business requirements. Regulations, privacy standards, and security requirements might preclude the "public" option is some cases. Should those entities fore go the cloud benefits because of a "naming issue"? Instead of names, let's focus on the real issues of a cloud - security, portability, interoperability, privacy, and trust.
* Originally posted on the ebizQ Cloud Forum on April 6, 2010.
Absolutely.
Let's not get hung up on names. The end game is to realize the benefits of the cloud business model. Yes, some options might be better at realizing a subset of benefits (such as cost reduction) than others but not all options are suitable for all business requirements. Regulations, privacy standards, and security requirements might preclude the "public" option is some cases. Should those entities fore go the cloud benefits because of a "naming issue"? Instead of names, let's focus on the real issues of a cloud - security, portability, interoperability, privacy, and trust.
* Originally posted on the ebizQ Cloud Forum on April 6, 2010.
Subscribe to:
Posts (Atom)