Conclusion

The marketecture around IP SANs and FC fabrics persists. Claims are repeatedly made by vendors in each camp about the superiority of their preferred protocol and topology that are not born out in reality. For example, nothing about IP storage makes it any more secure or manageable than FC storage. Moreover, neither IP storage nor FC storage truly breaks with the SCSI protocol or with the concept of a LUN that derives from the days of the parallel SCSI bus. Neither protocol fixes the problems created by the proprietary controller architectures of array vendors . Neither furthers in any significant way the creation of a common storage model. Neither delivers, on its own, the intelligence required to provision storage automatically to applications that need it. And, neither IP storage nor FC storage solve the ultimate conundrum : how to make data self-describing so that management tools can be created to realize not only capacity allocation efficiency, but also capacity utilization efficiency.

A final note about IP storage: It is a more elegant solution than FC if only because both the data path and the control path are handled through the same wire. In the past, elegance counted for something. In today's cost-constrained IT environments, it just might count for something again.