Well, one of the familiar things that popped up again at SNW was various vendors talking about FCoE -- Fibre Channel over Ethernet.
I really don't understand what the "controversy" might be about.
Must be a slow news day for some people ;-)
I've Written About This Before ...
... so if you want, check out the back story here.
For me, it's pretty simple.
- iSCSI has stalled in any enterprise that started with FC. We can debate about the reasons why this is, but the data (according to IDC) is pretty much irrefutable. One of my more controversial posts on this topic can be found here.
Any large enterprise guys out there want to talk about their successful large-scale iSCSI implementation? Anyone at all? Hello?
- At the same time, no one can argue with the economics of ethernet. For every FC chip built, there must be a million ethernet chips built. And, if you are familiar with the silicon business, it's a tough argument to fight -- massive scale leads to massive economies.
So, What's The Deal with FCoE?
Simple: how do you get the behavioral model of FC, and the economics of ethernet?
Throw away IP and the protocol stack, and run FC protocols in pretty much a bare-wire fashion on good, honest 10Gb ethernet.
Customers win big.
The only people who don't win here are vendors who can't afford the transition to this protocol.
So, Where's EMC Coming From?
We have a nice FC business. We have a nice iSCSI business. We have a nice NAS business, and CAS as well.
More or less, over the years, we've been about as protocol agnostic as any storage vendor can be. FICON, ESCON not to mention my old friend bus-and-tag. Several flavors of good, old-fashioned SCSI as well.
We're not especially religious when it comes to storage interconnects.
But we can spot a potential winner when we see one.
Let's face it -- it's not here yet.
We need more players shipping GA products. We all need to get comfortable with the deployability of the technology. The management tools will have to learn to understand the (potentially) different wire. And so on.
It's not up to vendors, or the press, or any consultant to decide whether FCoE makes sense.
As with anything else, the marketplace votes with their checkbook.
But make no mistake, I wouldn't be betting against FCoE ...