Organizations going down the private cloud path have some tough decisions to make. As highlighted in a new GigaOM Pro research report (sub req’d), “The Evolution of the Private Cloud,” most cloud management solutions are merely works in progress at this point, leaving customers with a catch-22-like situation. They know they need to get their cloud plans underway, but they also know that choosing one vendor over another will mean making functionality trade-offs.
On the one hand, there is VMware, with its relative openness in terms of hardware configurations, development frameworks and application types. On the other, there is Oracle, with its vertically integrated approach mandating, essentially, across-the-board Oracle. Microsoft hovers somewhere around the sternum, allowing openness in some areas, but keeping strong ties to Windows Server and .NET. Then there are traditional systems management vendors like BMC, CA, HP and IBM, all of which are taking different approaches toward third-party management of entire infrastructures. While George’s report stops there, the real-world cloud hype doesn’t, as cloud-focused startups like Cloud.com, Joyent and Elastra (and this list is by no means exhaustive) are touting their own cloud-from-the-ground-up management platforms.
However, it appears that many shops might just opt to stick with the vendors they know (such as Microsoft, Oracle and/or VMware), and use of Big Four systems management solutions will remain within the walls of large enterprises. This raises the question of whether cloud computing is indeed just an evolution rather than the revolution that some would have us believe it is. After all, if IT consumers are content sticking with their current vendors as both parties move prudently toward the cloud, that doesn’t indicate a dire need to obliterate the status quo. If current inefficiencies are as crippling as advertised, IT departments large and small would be jumping over to new vendors promising to do away with all of the dependencies and bureaucracies that make dynamic management and provisioning so difficult. Right?
The reality is that although moving workloads to the public cloud is in many ways a more revolutionary step than running workloads in a private cloud, the latter possibly presents more risks. Buying servers from Amazon Web Services is a low-cost, low-commitment transaction, whereas investing in hardware and software to power your internal cloud certainly means lots of money, and might mean an infrastructural overhaul. Plus, the market is quite volatile right now.
What if I invest in VMware today, and tomorrow Microsoft buys Citrix? What if I build a data center full of HP servers and Cisco switches only to see Cisco enter the server market and cut ties with HP? What if I bet my application environment on a startup (or Sun Microsystems) only to see it go bust, or (worse) get gobbled up by Oracle? Especially at a large scale, private clouds present many questions to which the answers aren’t necessarily clear.
To learn more about what’s driving private cloud efforts for both customers and vendors, read the full report. Also, be sure to attend Structure 2010, where vendors large and small will be discussing their visions of cloud management and what the data centers of the future will need to look like.