FCoE Or iSCSI, Does It Really Matter?

There is a lot of debate about Fibre Channel over Ethernet and converged network adapters. A CNA is a 10-GbE network interface card that supports multiple data networking protocols, basically TCP/IP traffic and storage networking. These adapters are going to support Fibre Channel over Ethernet (FCoE). The plan is to reduce networking cost of ownership by converging the data and storage networks onto a single adapter, which results in lower adapter, cabling, switch, power, and cooling costs.Why not use iSCSI? The game plan is that FCoE will use the current Fibre Channel protocol, preserving the existing investment in Fibre. It also eliminates the TCP/IP stack that can inhibit performance. Basically, FCoE is being positioned as an enterprise protocol and iSCSI as an SMB protocol. The result is another heated debate that, in the end, won't matter much. Here are some of the key points.

The FCoE guys will say that iSCSI has struggled to gain enterprise acceptance. Really? I've seen some very large shops using iSCSI quite successfully. One of the limiters to iSCSI is supposed to be that the storage guys and the network guys don't seem to get along and with Fibre the storage guys have their own cabling infrastructure. Won't FCoE have the same problem that iSCSI has now? You are going to converge on to a single wire, an Ethernet wire that will be controlled by the network guys, I would assume. FCoE is going to count on jumbo frames to make the transport work for FCoE -- that is going to take configuration work from the network guys.

The iSCSI guys will point out that FCoE is substantially more expensive at a per-adapter basis and that this is part of a strategy by those invested in Fibre Channel to keep customers using Fibre Channel. I don't know if I buy into that strategy, either, because the guys making the accusations are mostly in the iSCSI-only camp. The FCoE guys are quick to point this out, too -- of course, they are FCoE guys.

I think the real answer is going to be that the users will select the protocol that makes the most sense for them. If they have a huge investment in a Fibre Channel architecture and they can see value in converging onto a single platform, then that is what they will do. If they don't and iSCSI remains the cheaper alternative while still offering acceptable performance, then those users will select iSCSI. I think FCoE's initial impact will be to stop the defection of some Fibre Channel-based environments to iSCSI; as for broad acceptance, time will tell. My fear is that in the end it will be another protocol that needs to be supported, further complicating storage.

What is interesting is that both FCoE suppliers and iSCSI suppliers are attributing much of their growth to the continued rollout of server virtualization and that none of them that I know of have the ability to offer quality of service at the NIC level. That is going to be critical as virtualized workloads begin to compete for one or two 10-GbE segments that are doing network and storage communications. More on this next entry.

Track us on Twitter: http://twitter.com/storageswiss.

Subscribe to our RSS feed.

George Crump is founder of Storage Switzerland, an analyst firm focused on the virtualization and storage marketplaces. It provides strategic consulting and analysis to storage users, suppliers, and integrators. An industry veteran of more than 25 years, Crump has held engineering and sales positions at various IT industry manufacturers and integrators. Prior to Storage Switzerland, he was CTO at one of the nation's largest integrators.