Mass Storage has seen significant technological advances through the decades, but only in recent years has it begun its process of melding into the overall networking fabric in the enterprise - becoming a more integral component of the overall IT/Networking System - creating network branches that are becoming the management domain of the network administration staff.

Binding Storage technologies increasingly tighter to the overall enterprise network system gets us yet another step closer to managing IT resources from a more abstract level or vantage - to wit, Policy Based systems where allocation and conceptual management of compute and Storage resources on the corporate network are driven by Business Policies. Like it or not folks, that's where such choices ought, and ultimately must, be made. This of course does not preclude the need for, or hasten the demise of the network administrator. It does, however, create a more robust layer of abstraction between the plumbing minutia, and the setting of the levers and dials, driven ultimately by Business Process rules.

While today we remain in an era where a multitude of protocol and electro-optical conversions, transcodings and transformations occur before a byte under an application's control, lands on a disk platter - perhaps several time-zones away (via Long-Haul Remote Replication) - we cannot escape the conclusion that the fewer of these "impendence mismatches" the better. Adding Storage as a more tightly integrated network component, or branch of network nodes, presents yet another set of impendence mismatches, which must be recognized, confronted and managed - from the Ethernet connection for Web Services (XML-RPC, SOAP) based management, for example, to the OC-192 link which remotely replicates your binary assets 3 time-zones away. Additionally, Storage gets its tentacles into the IP space as a IP SAN component via GbE and presents yet further management complexities.

Advanced, leading edge Storage Systems themselves consist of several networks, loops (potentially), and protocols, and understanding the data ebbs and flows is central to making wise choices when planning deployments. From the System Area Network (around which all of your servers are built - not to be confused with Storage Area Network) which consists of the computer RAM, busses and I/O subsystem, to Ethernet networks, Optical Carrier networks (e.g., OC-48), Fibre Channel Fabrics, Fiber Channel Arbitrated Loops (FC-AL), etc.. Plug your IP network into the Ethernet port (for SNMP, XML-RPC/SOAP, Browser-based Management, etc.). Additionally, the GRID Computing model with its distributed nature creates an almost Fractal topology, introducing potentially a multitude of interconnects (for the early adopters) and long-haul taps for geographically dispersed GRID cluster linkage (these range from Packet over SONET (OC-3 .. OC-768), to DWDM over dark fiber, to wireless technologies and others). There can come a point where keeping-up with an enterprises' IT requirements to properly operate the business, and the management of those myriad of network systems rub - hard! A synergy is needed.

This paper endeavors to introduce and begin to confront the complexities of this convergence between what we know of Network Management today and Storage Management, and concludes with the open question of how this increasingly more tightly integrated set of technologies may work best as a streamlined System in the not too distant future.