Cloud’s Iron Fist
August 26, 2011

Dennis Drogseth

Share this

It may sound odd to combine “iron fist” with “cloud” – albeit it’s not the first time “iron” has found a linkage to cloud – e.g. IBM’s acquisition of Cast Iron last year. Clouds, of course, float above the earth and are generally rendered as white puffy expressions of nature’s whimsy (or else stormy expressions of nature’s revenge) when not directed at the high tech space. And for those of us who are old enough to count — “cloud” has been around in high tech circles for years as the all-purpose symbol for Wide Area Networks where, I might add, they generally retained the more whimsical, white puffy shapes of friendly clouds in rough architectural schematics.

But after spending more time than I might have liked relating just about everything I cover to cloud computing adoption – from CMDB/CMS systems, to cross-domain automation technologies, to BSM dashboards with advanced analytics, User Experience Management, to a service-centric vision of asset management, etc. – I have come to see cloud computing also has more than a few dark sides quite apart from the most obvious first-blush security and data control issues.

This isn’t to say that cloud computing is bad – in my own way I’ve come to appreciate it as a necessary transformative catalyst, though at times I feel a little like Dr. Strangelove in the process. And of course “cloud computing” as I always have to clarify – is far from monolithic. It is a motley (logically speaking) assortment of technologies and service choices that got thrown together in what I like to believe is secretly the occult meaning of cloud — “a foggy bank of new and old service delivery options optimized to capture media attention and support a lot of inbred vendor branding but with nonetheless real relevance to IT.”

In general, cloud is seen as expanding IT options for creating and deploying services, saving IT significant dollars in capex and opex expense, driving (forcing) IT and business clients closer together in dialog and planning, and even providing greater resiliency for delivering services and hence actually improving service quality. And in fact all this is borne out in EMA research. Cloud computing’s potential value is very real.

But in contrast to this expansive picture, I would like to devote this space to showing a few examples of how cloud computing can be constricting. How IT organizations that naively think cloud is in itself a cure for world hunger may find themselves taking a “journey” that’s more like the tragic Children’s Crusade than a Thanksgiving Day parade down Fifth Avenue.

Expectations: Cloud computing is long past the point of being purely a tech-centric term. It’s not only reached the ears of virtually all IT customers (internal/ external/ partner – you name it). It’s even been in prime time advertising on TV. All this, combined with an excessive and in my opinion often misleading array of marketing initiatives, is bound to set expectations too high, or at minimum askew. EMA research showing that 70% of cloud initiatives had to be redirected is in part a testament to this. CIOs and other IT executives caught in this vice between vaulting expectations, limited resources and a confusing array of options may easily find themselves choking for air.

Standardization: I think it’s quite interesting that a technology vortex that pretends to offer seemingly unlimited options also preaches the need for standardization and gets away with it. Certainly not all standardization is bad. And in fact consolidating and reconciling capex assets, management brand investments, and application portfolios is very much still needed. But many cloud companies would like you to standardize around their solutions for VM creation and the management thereof, or storage HW and SW, or at times even networking capabilities (HW and SW). Some “cloud” application provisioning solutions also require dedicated infrastructure. And this type of standardization, while philosophically OK, is historically just not how IT organizations evolve. Moreover, one could argue that if all these companies got their wish, big fish would soon swallow all the little fish and the natural competitiveness of market forces would become seriously degraded.

Cookie Cutter Services: In the mode of “standardization” – services themselves can become more cookie cutter like: predefined, prepackaged, and ready to roll like a McDonald's cheeseburger on an assembly line. All this is very good if you want fast food. But not all application or other services fit the fast food mode, and certainly not all consumers thrive on a steady diet of manufactured, rather than carefully prepared, delights.

IT Organizations Caught in an Iron Vice: Cloud computing is disruptive to conventional IT organizational models. From a top down service management perspective, this is arguably a good thing – forcing a long needed acceleration towards more cross-domain processes, dialog, management technologies, and even organizational models. However in a techno-centric market, there is little to nothing to address this requirement directly associated with cloud adoption. Moreover, cloud can create new silos – virtualization experts – that may run the risk of adding to IT’s feudal kingdoms if they don’t report into a true, cross-domain service management team.

Service Provider Lock-In: All cloud service providers are – guess what, “service providers.” And what do I hear most when I ask IT professionals about their experience with service providers? “Rigidity” sums it up nicely. Service providers will offer you self-protective contracts reviewed by lawyers who know quite a bit about self protection and which are almost invariably skewed to their interests first and yours second. Now, this is not to say – regarding cloud or any other service provider (I depend upon quite a few myself) that all service providers are bad. But the notion of achieving “freedom” through the use of “service providers” quickly becomes an oxymoron unless you know exactly what you’re doing. Expanding on that last sentence would require a blog in itself, so I won’t attempt it here.

As most of you probably remember, the vagaries of cloud service provider performance were recently dramatized with all the buzz surrounding Amazon’s April reports of connectivity problems impacting its Relational Database Service that covers a broad area along the U.S. East Coast. And Compuware data indicates that North American companies estimate organizational/business losses of about $1 million a year from degraded performance from cloud-based applications. On the other hand, some service providers, and my favorite example is Akamai, have actually offered sophisticated user experience monitoring to their customers to help ascertain where and how their value is best achieved. This stands in contrast to most (all?) SaaS providers who don’t even seem to know what user experience is, but stick to internally defined metrics for availability.

I won’t pretend that this list is complete and welcome any additions or added thoughts from you. Nor am I suggesting that we all try to go back to a pre-cloud world (although I might enjoy writing more about other things for a change). What I am suggesting is that cloud – in spite of the white-puff appeal of its name – can also be dark and stormy if it’s not, as the now old song goes, seen from both sides.

Related Links:

Dennis Drogseth is VP at Enterprise Management Associates (EMA)
Share this