As organizations continue to grapple with managing unconstrained storage growth, they are repeatedly working to improve efficiency and reduce costs largely through greater consolidation in hope that this will improve utilisation. However, there often exists a fundamental disconnect between how storage technology and services are delivered and the way in which they are funded and accounted for.

We shouldn't forget that the basic premise for adopting networked storage technologies and implementing SAN infrastructures is that storage in a direct-attached configuration is inflexible and inefficient. By networking, we are better able to allocate and share storage where it is truly needed. At least that is what we believe and fundamentally what should be true.

Yet in many large organisations that have embraced SAN-based storage, utilisation metrics continue to be far below expected levels. One reason may be that the purchasing and funding mechanism for storage has not advanced concurrently with the technology. Storage equipment is still largely purchased as a capital asset associated with a specific project, such as a new application rollout, and as a result, it is still treated as an asset that is 'owned' by the group providing the funding rather than a shared resource.

While it may have made sense with silo'ed application-server-storage stacks, this approach does not jibe with the concept of storage as a shared network resource. It simply runs counter to the capabilities and best use of the technology and desired efficiency goals. A storage array or tape library in a networked storage model should not be 'owned' by a business unit. The mechanism for funding storage for new projects as well as the ongoing chargeback or billing function must change to align with the new usage models.

This ought to be a solvable problem. Corporations already have shared IT resources such as datacentres, networks, and even mainframes that are funded across the organization. An application project doesn't typically build its own network or data center -- it shares these resources, and capacity is expanded as needed. The fiefdoms of isolated SAN islands and dedicated arrays needs to evolve toward in true shared infrastructure and financially, it needs to be handled that way as well.

Jim Damoulakis is chief technology officer of GlassHouse Technologies Inc., a leading provider of independent storage services. He can be reached at [email protected]