Last week I had an interesting conversation with a US State client, who runs a central organization providing infrastructure services to some State agencies. The IT organization there is quite decentralized (or – al most –federated) and each agency can run its own IT infrastructure, although quite a few have their servers hosted by the central organization.
The conversation was about how the central organization could leverage cloud computing to reduce costs, especially in the short term. From what the client told me, their data center is already virtualized so they could already provide infrastructure as a service to client agencies (although we did not go deep enough to fully prove his claims in terms of SLAs, pricing models, and so forth). Additional data point: their data center and related services are heavily insourced (i.e. there is a prevalence of government staff plus a few contractor working on a time & material basis).
I soon understood that his interest in cloud computing (and related saving) was about structuring their own services and not about exploring either the possibility of relying on external providers to meet some of their needs or to establish themselves as a “storefront”, should agencies be willing to move some of their own workloads to external on-demand resources (cloud infrastructures or IT utilities).
Like many who are heads down in their consolidation and virtualization efforts, he was overlooking that cloud computing can be as much an ally as an enemy of those centralized efforts (see previous post) and that, when looking for fast cost saving opportunities, some of these may be seized by individual agencies that source some of their infrastructure or software requirements to a “public cloud” service provider.
Looking at cloud computing as one out many different alternative sourcing and delivery models both for individual agencies and central organization tasked with consolidation efforts, and balancing pros and cons from an overall budgetary perspective is the way to go.