Tuesday, March 2, 2010
Of Health Care Bills and Hybrid Clouds...
Wholesale conversion of massive data centers laden with legacy systems -- including highly efficient mainframes -- seems impractical and unlikely. Certainly, it is easy to overlook this somewhat pedestrian issue. The lure of getting up and running faster with new products and services is as compelling as the ability to scale on-demand to handle unexpected loads. A compelling argument could be made that these safeguards alone justify the change, if for no other reason than to prevent lost business due to decreased performance or downtime. But cloud management issues, particularly security and compliance, are not sufficiently resolved yet to enable CEOs, CIOs and boards of directors to endorse a wholesale move to the cloud.
What seems most logical is an evolutionary process through stages of the hybrid model. This likely starts with a shift of internal virtualized data centers to deploying external clouds for peak demands to meet the needs of internal users.
As an enterprise gets familiar and comfortable with an external cloud model, it will more readily evolve from peak demand to on-going hybrid, where IT departments rely on external cloud providers for continuing support of non-core functions.
This evolution holds endless possibilities, but we wonder how quickly it can happen. Looming large in the decision process is the "S" word. It is reasonable to expect customers will require the same level of management and security (policies, systems and processes) in a hybrid cloud that they have with internal data centers or private clouds. Today, customers perceive risk -- what happens if my data is co-mingled with others when I go off-premise? Vendors have not addressed this perception to the extent that customers are ready to move.
At this point, it isn’t clear whether the health care bill or full-scale cloud adoption will evolve first. They both hold enormous promise for change, but both seem to be progressing in fits and starts with non-negotiable issues still unresolved. In a world in which success seems to be increasingly measured by wholesale change rather than incremental development, we're not yet ready to bet the farm on which gets resolved first.