Cloud CIO: The Next Generation Cloud Offering
Of course, it's early yet. Much of the initial commentary has evaluated this initiative as a threat to Amazon or other cloud providers. The "who wins, who loses" approach is tempting, but, really, in such a rapidly growing field, CSPs should be focused on how to ease adoption, not who is top dog, which means they'll welcome anything that reduces barriers to use. In any case, CSPs make money by selling computing service, and anything that eases consumption is a positive for them. A more telling question is to what extent CloudFoundry will build a community and ecosystem surrounding the technology offering. And beyond the typical issues of community building, one would think two critical constituencies will have questions about their potential involvement.
For CSPs, more telling than which PaaS is the more powerful is likely to be a potential reluctance to commit to an offering from a vendor who is a key supplier to a competitor. This will probably be even more true for non-VMware based CSPs. There could be a natural concern that information shared during a discussion about business strategy with the CloudFoundry team might find its way to the ears of the unit that supplies software to other CSPs.
For end user organizations, a natural concern would be the depth of commitment by VMware to CloudFoundry. The software product landscape is littered with the abandoned open source initiatives of vendors, who launched an open source product with great fanfare and then lost enthusiasm once they realized how long it takes and how much work is required to build a vibrant community. This abandonment isn't the result of lack of desire on the part of the vendors upon initial launch; it just reflects the fact that later, when it comes time to invest in the product's growth, higher priority items displace the money or headcount required for the open source initiative, and one day it becomes evident that the product has grown stale and the community has petered out. Obviously, an end user is going to evaluate the prospects for an open source product's longevity very carefully before jumping on board.
These reservations are not to criticize the CloudFoundry initiative itself. I hope Ive made clear how important and necessary such an offering is. The fact that a proprietary software vendor has had to step up to make this happen reflects, in my view, the paucity of imagination on the part of the venture capital community. The need for high-productivity, application development streamlining is obvious - why wasn't CloudFoundry or an equivalent initiative funded as a startup?
For a cloud CIO, the takeaways from these announcements are the following:
Sticking a cloud infrastructure into an cost-uncompetitive data center (and that includes operational inefficiency as well as kit inefficiency) consigns you to a high-cost provider position, not a comfortable place to be in the future. 2. Accelerate your plans to move to cloud computing. A two-year plan to stand up a dev/test cloud is just too long a timeframe. It risks solving yesterday's problems and being bypassed by tomorrow's solution. 3. Consider where to invest your staff skill development dollars. If an external company can operate a high-efficiency PaaS that your applications can use, perhaps you would be better served by leveraging that and directing your staff training and recruitment toward applications, skilling up on people who can place business logic on top of the PaaS infrastructure. 4. Ruthlessly standardize offerings. I made this recommendation last week and make it again even more strongly in the light of these new developments. "Have it your way" is a great restaurant jingle, but it's hopelessly expensive in IT. 5. Consider using differential pricing to guide business unit behavior. Many business units will insist on "having it their way," and in the absence of any reason not to, will expect IT to deliver custom configurations, etc. Offering the standard configuration at one price and the custom configuration at, say, 3X the standard price sends a price signal and allows the BU to evaluate how important the custom configuration is (hint: at a higher price, it may not be nearly as important).
This is, perhaps, the most interesting time in IT since the rise of the PC. Certainly the democratization and explosion of applications enabled by cloud computing resembles that heady period, when it seemed like new developments sprang forth daily. It's fantastic to see companies like Facebook and VMware contribute to this innovation and provides anticipation for new developments that are on the horizon.
Bernard Golden is CEO of consulting firm HyperStratus, which specializes in virtualization, cloud computing and related issues. He is also the author of "Virtualization for Dummies," the best-selling book on virtualization to date.
Read more about virtualization in CIO's Virtualization Drilldown.