Which technology will enable networked storage?
A debate has been raging on Wall St. over which technology will enable networked storage: Will the backbone be Fibre Channel (FC) or is it Ethernet with Internet Protocol (IP)?
As often happens, this debate isn't as prevalent on Main Street, where data suggests that customers are using Fibre Channel at an accelerating rate to implement storage area networks, while sales of network-attached storage is booming. As I wrote in a previous column, the explosive growth of data is creating increased demand for storing and managing all the data.
Nevertheless, it is important for the investor to understand this debate and monitor new developments in technology that may affect the outcome.
But first let me be clear: I view Fibre Channel as a unique significant enabling technology that will continue to grow in usage over the next 3-5 years--which is as long as any investor's time frame should be. This is not to say that I don't see the elegance behind the argument that all communication is moving to Internet Protocol, and the cost/performance advantages of a unified network are great (in other words, separate LAN and SAN technologies should converge over time).
The idea is to develop a high-speed interconnect between storage and servers that improves the speed and flexibility of the current standard SCSI (small computer system interface) protocol. Fibre Channel was developed with just that in mind, and providers such as Qlogic, Brocade, Ancor and Gadzooks (to name just a few) offer products that provide the storage network infrastructure based around this technology.
It's not that Ethernet can't get better at transferring storage, it's just that Fibre Channel is the best that exists today. Remember, transferring storage blocks has different functionality needs than transferring Web pages, emails or files across a LAN.
New methods of interconnecting storage and servers are required as soon as possible. While the standards for wide-scale Fibre Channel fabrics are still being sorted out, and interoperability between suppliers remains difficult, the technology does work and is being launched. Standards for SCSI over IP, and required interface cards, switches and additional software drivers/layers do not exist yet; they will be available over the next few years, not months.
Finally, interoperability with existing storage is a prerequisite for any storage network solution, and testing takes more time than anything else. Fibre Channel's years of testing gives it a substantial lead in this regard.
As far as I can tell, only the most rudimentary technology works today. Some companies have announced solutions for the future, but few are specific as to when these solutions will be launched. As stated earlier, I have no doubt that this technology will emerge. However, I think it is 4-5 years away, given all the work in front of companies focused on this area.
Meanwhile, I believe the Fibre Channel companies will evolve and even embrace the benefits of emerging storage over Ethernet/IP--much like the early NAS players, such as Network Appliance, which has embraced beneficial features of the SAN.
Bottom line: It's too early to compare technology that doesn't exist with technology that works today. And I don't think the new technology will be disruptive to today's leading players.
So, watch for upcoming announcements from both camps. But remember storage platforms and especially storage networks will be fundamental building blocks of the Internet infrastructure. As such, the winning players should be part of every investor's technology portfolio. For the investor it's not SAN vs. NAS vs. Ethernet/IP--invest in what is available and being launched right now.
Philip Rueppel is a research analyst with Deutsche Banc Alex. Brown. Rueppel's comments that appear herein are not a publication of Deutsche Banc Alex. Brown and may not represent his complete or current opinion with respect to any company. Persons who want to make an investment decision with respect to any company mentioned by Rueppel should obtain a copy of his current and complete opinion as contained in the most recent publication of Deutsche Banc Alex. Brown. Rueppel's opinions are not intended as an offer or solicitation, nor as the basis for any contract for the purchase or sale of any security, loan or other instrument.