FCoE Is Nigh, So Where Are the Cheers?

Arthur Cole

Fibre Channel over Ethernet (FCoE) is finally entering commercial introduction after nearly a two-year ratification process. So why are doubts about the efficacy of the technology being voiced now by those who stand to benefit from its adoption?


I can't say that anyone is getting cold feet about FCoE, but over the past few weeks or so, there's been a growing chorus of experts either explicitly highlighting the challenges in implementing the technology or pulling back from what should be an aggressive rollout over the next year.


Exhibit A is this report on InfoStor that has Emulex' Joe Gervais talking mostly about the disruptive nature of FCoE and its fairly complex implementation challenges. The magazine quotes Gervais saying:

"The challenge is bringing two organizations--the networking team and the storage team--together, which in the past, have been typically siloed. We're talking convergence, so it's going to take a lot more cooperation than what we've seen in the past with these two groups. Convergence isn't hinged on technology, but on a mindset of organizations coming together."

He also points out that installing FCoE will likely require coordination among numerous vendors -- namely storage, adapter and switch providers -- which is not exactly music to the ears of those already dealing with multi-vendor environments.


None of these concerns are deal-breakers, mind you, and the article does spell out the substantial benefits of FCoE -- less cabling, fewer network devices, lower power -- but for a firm banking on FCoE for much of its future livelihood, talking up the negatives at this stage seems a little odd.


Next we have NetApp, which is playing both sides of the iSCSI/Fibre Channel fence. Co-founder Dave Hitz has overcome much of his initial skepticism over FCoE, but he can't help but wonder about the timing. At the moment, anyone thinking about an FC upgrade would do better to wait a year or two to see how FCoE shakes out. But if you want Ethernet-based storage, then why wait? An iSCSI NAS is capable of meeting just about all storage needs and you can still fold in your FC storage when the time is right.


Brocade, meanwhile, is lowering expectations for a major FCoE push in 2009, with GM Ian White saying recently that the company will introduce products in the first half with an eye toward ramping up sales in 2010. Under the company's FCoE roadmap, initial products will be an FCoE switch and a converged network adapter (CNA), with a storage OEM agreement for a tier-one HBA by late 2009.


Still, for those chomping at the bit for FCoE, a handful of systems is making its way to the market. QLogic has begun shipping the 8000 series CNA, packaging it with the Cisco Nexus 5000 switch. The device is available with or without optical transceivers depending on whether you have copper or fiber networks.


There's no doubt that FCoE retains substantial support both among vendors and users, many of whom would hate to see their installed Fibre Channel SANs get outclassed by competing Ethernet technologies. But with the technology on the launch pad, it's surprising that there's not a bit more enthusiasm out there.

Add Comment      Leave a comment on this blog post
Aug 22, 2008 9:39 AM Stephan Stelter Stephan Stelter  says:
Interesting article! I am a bit surprised that you are surprised with the level of enthusiasm. I think few customers who lived through the wonderful compatibility mayhem around 1 Gb FC are jumping for joy around the prospect of the same with FCoE. Customers probably can't wait to figure out which version of BIOS/firmware on their new-fangled CNA works with which version of firmware on their new-fanlged Nexus 5000. I wonder if Cisco support will help with that or if you'll have to weed through telephone menus and supervisor escalations to get to a Nexus engineer?I'm sure the FCoE vendors are equally excited about the prospect of getting JNI'ed or McData'ed (this expression sounded much better coming from Scott McNealy in his Chicago Comdex keynote, when he said "no one wants to get Wang'ed or DEC'ed"). Compatibility problems plague new product introductions and I'm betting that FCoE will be particularly painful. If I'm Emulex or Qlogic and I'm courting the usual suspects (HP, IBM, Dell), I would want to make sure their demands won't quickly drain my support resources.Perhaps those FCoE vendors with traditional FC products see the short-term consequences of FCoE exuberance. Why should customers consider buying traditional FC products (the only products these vendors can sell in any volume today) if everyone is enthusiastic about FCoE?I'm also a bit puzzled by the idea that users should be passionate about their installed fibre-channel hardware getting outclassed by pure-play iSCSI over Ethernet. Managing volumes, capacity utilization, and application I/O performance will be a different skillset than managing ports, packets, and routing tables for the foreseeable future. If the tools to do my job improve, why would I chose to use my current, less-effective tools? Yes, a screwdriver removes a screw, but a power driver does the same job faster. Reply

Post a comment





(Maximum characters: 1200). You have 1200 characters left.



Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.