Storage Cycles

I recently read an article written by Chris Mellor, under The Register banner, titled “iSCSI- Game Over”. Entry can be found at: http://www.theregister.co.uk/2008/10/15/fcoe_io_kill_iscsi/. The article suggested, at least from a supply-side perspective, that iSCSI is dead and that Dell paid entirely too much for EqualLogic given this demise. Well, given the shape of the economy and acquisitions in general, perhaps Dell did pay too much. But I think that this has more to do with broader financial market issues than the intrinsic value that EqualLogic brings to Dell. When I say “supply side”, what I’m suggesting is that the article factored in input from suppliers with a keen interest in Fibre Channel, not end-users or suppliers that firmly believe in the merits of iSCSI. I’d like to believe that markets are largely shaped by users and less by suppliers. Especially today when information is so readily available and users often know so much more than suppliers about how things really work. All suppliers have interest in prolonging what is good for them. The danger for entrenched suppliers is that the “Innovator’s Dilemma” sets in and established vendors are over taken by the risk takers, the innovators; those that really think outside of the box in delivering value. A unified network built around Ethernet is the way to go, but to assume that FCoE will become the standard means for networked storage connectivity, edging out TCP and future innovations in both iSCSI and Enhanced Ethernet, which may well benefit iSCSI and the overlaying applications, is frankly a bit premature.

I know Chris Mellor and have found him to be a very fair and detailed in his reporting. I think that the article deserves more discussion; something that I know Chris would be more than welcome to entertain. His article was not intended to be a conclusion but rather an introduction to constructive discussion. One of the suppliers referenced in his article is a company that supplies Fibre Channel HBAs. The company does a great deal more than Fibre Channel HBAs, but a substantial portion of its revenue comes from this source. As part of its mix the company sells iSCSI HBAs. In fact, up until about a year ago, the company was quite vocal that iSCSI was very strategic to its future. That was until FCoE became a priority, not because any End User Council or strategic end-user suggested that it become a priority. In fact, I seriously, with all due respect, doubt that this company has a great many end user relationships. No, FCoE became a priority, I believe, when certain OEMs whose interest is to prolong Fibre Channel from a business perspective told them to make FCoE a priority.  The dirty little secret was that not succumbing to the demands of its OEM customer clearly meant that its Fibre Channel HBA business “could be” shifted to its competitor(s).  Now, that’s the free market at work in the fibre channel supply chain. Establish two strong vendors and have them duke it out and apply pressure and threaten, however subtle, to shift share for lack of conformance. I kind of feel sorry for the executive quoted in Chris’ article. The role he plays reminds me of Charlie Durning’s role as the Governor in “The Best Little Whorehouse in Texas”.  He’s got some pretty tough constituents. I personally witnessed the old side step, in action. God help him if he had expressed confidence in iSCSI. The speed at which this company shifted focus from iSCSI to FCoE was simply amazing. It had nothing to do with end customer revelations.