To: Bob Frasca who wrote (9573 ) 3/11/2000 3:48:00 PM From: D. Newberry Respond to of 17183
Hi Bob, << What evidence is there to indicate that NAS is going to supplant SAN in the storage world other than some highbrow white paper? >> Your questions are the same I have been concerned about for some time. I have been long EMC for over a year, and accordingly rewarded. I have been doing considerable research on the SAN vs NAS debate recently, and have yet to reach a definitive conclusion. I am staying with EMC for the time being, however, and will continue to monitor the situation My background is in Network Engineering, but I have no direct involvement in storage. My analysis is therefore second hand. However, I think it is useful to look at this business from two perspectives, the software as well as the hardware. In looking at the hardware components of NAS vs SAN, I am inclined to give NAS the edge. After all, they use standard Ethernet technology which is cheap and standard. 10G Ethernet is just around the corner and it will work very well. SAN, on the other hand, is a standard specific to storage. I would speculate that the costs will be higher for this hardware simply because the development is spread out over a much smaller base. The ultimate transmissions speeds will also max at 2G. The area I stumble on is the software element. If you are to assume that NAS (ie NTAP type devices) will win this contest, I believe you are assuming that storage will be a fairly non-complex commodity product in the future. An NTAP device, as I understand it, is easy to plug and bring on line. My experience in networks is that you generally trade simplicity for functionality. My assumption therefore is that the NAS product, ie NTAP, does not have the functionality that the traditional EMC product line has, based on its' simplicity (granted that is a big assumption to make). Now, taking that thought a step farther, I have a hard time believing that future storage requirements will be less complex than they are today. If you assume that our internet centric future will be centered more on storage and retrieval rather than mere processing, then I also assume that software management of the storage components will be extremely important -- at least in the larger data centers. With that line of reasoning I have decided to stay with EMC. They clearly have the software, not to mention great customer service. They are moving into NAS, and if Ethernet hardware becomes the preferred transport of the future, I would think EMC can make that transition easy enough. Their software component, however, cannot be easily duplicated by other vendors. I believe that is the competitive advantage that EMC has. Again, I have no direct experience with large scale storage systems. I would appreciate feedback from any posters that are in this field. It would be nice to get some direct insight. By the way, I think Gilder is great at the high level long term view of the universe. IMO his recent move into the stock picking world is misplaced, so I really don't put any stock in his NAS vs SAN arguments. This is the same guy that stated the other day that Cisco was doomed. Not that Cisco is guaranteed success, but his Sonet arguments made no sense from a technical perspective. Regards, DN