I was looking over some research at Ovum that discussed the concept of the “local cloud.” The initial take is that local and the cloud are not necessarily terms you would put together.
But I would argue the opposite. In a truly open cloud, there are various flavors. The public clouds will be far-reaching and myriad in terms of what runs on them – with providers of this type of cloud being huge providers like Amazon, Rackspace, IBM, Sun and even telcos. Vendor clouds will be highly specialized – take Sugar On-Demand as an example. Partner clouds will be a mix, offering vendor driven apps like SugarCRM, but perhaps with custom extensions or other open source or other functionality on top of the cloud-based offering.
And then there are private clouds – these are cloud-based environments dedicated to a single end user organization. For example: a large enterprise serving multiple locations via a single datacenter.
Just because the cloud and the Internet seem to be interchangeable does not also mean that the cloud has to be a “global” concept. Just as Sugar On-Site deployments leverage the web for delivery of CRM functionality to end users – private or local cloud deployments need not span the globe or be provided by vendors or partners seeking peak scalability.
To get back to the Ovum point – I would also argue that the “local cloud” is a hybrid of sorts between the private cloud I describe above, and the partner cloud. Both of these flavors have some level of scalability in mind – but the amount of dedicated resources per deployment, and localization/verticalization inherent in these models versus vendor or public cloud flavors – means that some level of “local” is employed here in the strategy.
All in all, I seriously like that many parties in the software value chain can benefit from cloud computing concepts. The true cloud opportunity is far greater than the lock-in minded, multi-tenant SaaS vendors would like you to believe…