Do you foresee NAS storage eventually being served from the SAN storage pool as the most effective storage management strategy (utilizing the extensive scalability facilities of a SAN pool and the backup/mirroring facilities with an appropriate storage management appliance)?
No, not necessarily. While a NAS head might seem like an obvious solution, there are just as likely to be other structures besides SAN on the backend. The recent closing of Cereva indicates that the NAS-head idea might not be as terrific as many think. There can be distributed NAS, clustered NAS, multi-layer NAS, all kinds of things where SANs don't necessarily come into play.
The key to all this is what type of data management benefits can be achieved. For the most part SANs have the potential to be strong in storage management and NAS has the potential to be strong in data management. The amount of leverage that can be exerted on DATA MANAGEMENT by either SAN or NAS will eventually become the leading solution type.
From my perspective, this can only work for SANs in the long run if storage management and data management can be integrated in a meaningful way. This is a tall order because it means much tighter integration between filing, storing and wiring than exists today. Not that it can't be done, but there is considerable heavy lifting and the critical mass of developers hasn't formed yet. Considering the history of interoperability in SANs, one has to wonder whether or not a standards-based coalition can even form.
But NAS is also a fragmented industry. Many vendors fail to understand the importance of data management in their core product designs and try to add it as a bolt on. Microsoft's strategy of keeping client-filing protocols proprietary also creates huge problems for everybody (vendors and customers alike). IT shops everywhere need to recognize that Microsoft is clearly not acting in their best interest on this issue and put pressure on them to open up their filing protocols for the rest of the industry. If they don't everybody will continue to pay excessive development and royalty tax for many years - while receiving sub-optimal data management capabilities.
So, there are problems on all fronts. That's what makes it so interesting. The business hurdles are harder to fix than the technology hurdles. My advice is not to believe that a future architecture is inevitable and to keep your options open.
Editor's note: Do you agree with this expert's response? If you have more to share, post it in one of our .bphAaR2qhqA^0@/searchstorage>discussion forums.
This was first published in July 2002