This blog has moved...

If you're visiting this blog via Blogger, please note that the WebLayers blog has been moved as of December 29, 2010. You can now find it at: http://blog.weblayers.com/

Is distributed governance a redundant phrase?

David Linthicum's Cloud Computing blog over at Infoworld makes solid points about SOA governance's move into the clouds. As I've addressed in my November 18 post, "Governance: a cloud computing strategy's silver lining," the key to building a seamless enterprise architecture is through governance.

The guiding principles of SOA and all out efforts to break down the silos we've inadvertently created over the years has led us to a point where governance can no longer be an afterthought as it is in so many instances. As more applications and services are opened up to a greater amount of users - many of whom may not have been anticipated when the original architecture was being sketched - there needs to be a more concerted focus on the way governance is introduced and carried throughout the software development life cycle.

This is most evident in a company's decision to solely focus on a run-time governance approach. Now there are certainly places where runtime governance is most relevant and necessary - to gain better control of the runtime environment - and when it's equally critical to enforce governance at the very beginning design stage. While each approach, runtime and design time, brings value to the architecture and are complementary to each other, what is becoming most relevant and critical as companies move to cloud computing is the distributed governance model.

To this end, WebLayers announced in July the industry's first fully distributed policy management platform. As more companies start to roll out products in this arena, I suspect that we may no longer be differentiating between distributed governance and non-distributed governance. It will simply be that distributed governance will be the de facto approach due to the fact that nearly all architectures will be distributed to some degree as we continue to extend our applications and services to other departments and divisions as well as external customers and partners.

Computer glitch delay your unemployment check? How to tell the kiddies Santa won't be coming

The latest computer glitch comes from the California Employment Development Department. Due to the system's inability to record extended unemployment benefits, over 100,000 residents have not received their checks and may not until after December 25.

Recognizing back in 2002 that the 30-year old computer systems were in need of a serious upgrade, the federal government gave the state of California $60 million to address the current and anticipated issues. While some progress has been made, it's clear that the project is not complete.

Of course, this isn't what unemployed residents want to hear when their bills are mounting and the holiday season is upon us.

So how do you explain to little Suzy or Bobby that Christmas may be delayed? Articulating the complexity behind the computer glitch may be a bit heady for some parents so here are five excuses to tide you over until the unemployment checks clear.

1. Santa decided that this year that he would split up his trip into two nights and you're on the list for the second delivery that is expected after the 25th, or somewhere thereabouts.

2. There was a layoff in Santa's workshop and with less elves working, they weren't able to make as many toys as they used to.

3. The reindeer caught the H1N1 virus and won't be able to fly until later this month when they're feeling better.

4. Santa's credit card was frozen in the North Pole and it won't thaw until the end of the month.

5. Your email to Santa got stuck in a spam filter.


While the California residents patiently wait to see if the glitch does in fact delay their checks for another few weeks, the issue calls into question the flexibility of the IT infrastructure. You have to wonder how much governance has been applied to the system design and why such a slight change such as the extension of benefits could trip up so many.

Toys R Us Computer Glitch Charges Customers Double

It was a Black Friday indeed for Toys R Us customers who discovered that they were charged double for their purchases due to a computer glitch.

Now customers have to straighten out the charges with Toys R Us as well as their banks to address the issue and the potential overdraft fees.

While Toys R Us did admit that the computer glitch happened at several stores, it didn't give the number of consumers that were affected. However, setting up a hotline gives me a pretty good indication that this isn't a one-off incident.

Where the IT infrastructure went awry is still not clear though I suspect the architects and developers are working harder than Santa's elves right now to fix the issue.

I empathize with them and wonder, like their customers, if this issue could have been avoided altogether. Was it an oversight in the software code or the inability to scale to meet the demands of the busiest shopping day of the year? Along those lines, it may have been the unanticipated backlash from a highly successful social media campaign. In case you missed it, Toys R Us is currently the fastest growing brand on Facebook with 400,000 fans. The company's aggressive Black Friday marketing campaign gave a sneak peek of Black Friday deals only to its fans. While the campaign was successful at driving customers to the stores, it is IT that will bring them back.

From the outside looking in, the company made smart technology choices and aligned with top vendors. Whether they have a service oriented architecture, cloud strategy or other best practices in place is unknown to me and now is certainly not the time for I told you so. However, you can guess that an IT audit wouldn't be an unreasonable request once the dust settles. Of course, I'd question the governance that's in place to prevent these types of glitches but that's only one piece of the puzzle.

I'm sure this will all be fixed by the time you read this though I still believe that an ounce of governance prevention is worth a pound of cure.

'Computer Malfunction' is IT's Answer to the Dog Ate My Homework

I came across this cartoon recently and thought you'd also find it amusing since 'computer malfunction' has become the catchall phrase for anything that's gone awry. And it's a convenient excuse for explaining outages, downtime or delays.

Though I have to wonder why we still accept this as a reasonable explanation as if the computer alone is the source of the problem.

Which brings me to today's first rhetorical question: if humans programmed the software and there's a malfunction, who is responsible, the human or the computer?

I'd venture a guess that once the source of the issue is identified, the human likely caused it.

If you've been following this blog, you know that I've ranted about computer glitches and their source for some time now. What continues to amaze me is the constant stream of computer glitches and the complacency of the industry and consumers.

Along with the recent FAA computer glitch that led to flight delays all over the U.S., there was the temporary search outage on eBay that happened just two days later. The steady stream of computer glitches continue to make headlines yet the underlying causes remain hidden.

Now in the spirit of rhetorical questions: has there been a week since the decade began when we didn't hear about a major computer glitch interrupting business in at least one pocket of the world?

Sure, there are technical explanations for the particular issue at hand but we seem to have lost sight of the fundamentals of software development that could have prevented or at least flagged these issues before they affected hundreds if not thousands of people.

Which brings me to my next rhetorical question: if we have the technology to build a more stabile technology foundation, why aren't we applying those principles to the development process?

It's the shortcuts that will inevitably lead to shortcomings that impact the consumer. This doesn't mean we have to add extra work to the development process. On the contrary, if we address these potential issues as the software is being designed, we can avoid many of the mistakes that will be surely be revealed at some point in the future.

Which brings me to my final rhetorical question of the day: why do we, as an industry, focus on addressing issues after the fact instead of preventing them?