From a storage network admin point of view, I couldn’t agree more. For example, Baker referred specifically to VMware, where iSCSI makes it possible to maintain the same touch points that you have in a real environment unchanged in a virtual machine. That includes the initiator-target link, running backups in the same machine, moving the guest environment to a different VM, or moving a VM to a different ESX server.
By contrast, to do the same things with an FC SAN you need to touch the storage network at several points, which may include an FC switch and one or more HBAs. Also, in a VMware ESX environment, the Hypervisor -- not the VM -- owns the HBAs, which means changing things sensibly. Ditto if you want to move back to a physical machine.
Is that a big deal? Probably not, as skilled admins can do that in minutes and with a hand tied behind their back, but it still takes some time. Moreover, depending on who calls the shots in your datacenter, letting the VMware admin manage VM related-storage (as you are forced to do with FC) could be an improvement rather than an obstacle. You may want to read what they say about this on the IT 2.0 blog.
So there you have it: Server virtualization is a double-edged sword that can cut on both the iSCSI and the FC side. Our great-grandchildren might still be debating which one is best 100 years from now.
Join me on The Storage Network with questions or comments.