Cisco Blogs


Cisco Blog > Data Center and Cloud

Unified Fabric – a Myth or a Reality?

April 8, 2008
at 12:00 pm PST

The topic of”a unified fabric” has recently been generating some buzz in the blogosphere, with both advocates and skeptics weighing in. I actually think the question of”Myth or Reality” has been answered. We currently have customers that use both InfiniBand and iSCSI to create a unified fabric and each technology has staunch supporters. So, I think we are past the question of “if” and on to the question of”when?”So. what are the current inhibitors for broad-based adoption of a unified fabric?Off the top of my head, I can think of a couple of points to consider:

  • Architectural Intertia:There needs to be some compelling reason to change--cost, performance, features, etc
  • Investment Protection: related to inertia, there needs to be a compelling reason to abandon infrastructure that is has been bought and paid for, especially if it is working well
  • Operational Characteristics: unified fabric is not just an exercise in payload encapsulation, but in maintaining the correct operational characteristics, such as Fibre Channel’s losslessness
  • Operations and Management Disruption: what is the learning curve for new technologies--how disruptive will they be to existing design and operations best practices

From an architecture level, I don’t think organizations will move to a unified fabric until they encounter a solution that delivers compelling answers in all these areas--but I think eventually they will move. That being said, when we get down to specific technologies, I don’t think there is any”one protocol to rule them all”. Fast forward a few years and I think your typical enterprise data center will support a mix of iSCIS, FCoE and good old Fibre Channel.So, what do you think--in the next three years, do you see yourself considering or deploying a unified data center fabric? Why or why not?

In an effort to keep conversations fresh, Cisco Blogs closes comments after 60 days. Please visit the Cisco Blogs hub page for the latest content.

4 Comments.


  1. On the subject of conversion to a Unified Fabric. I attended the IDC Data Center and Virtualization Forums this week in San Francisco, and sat in on a presentation by a FC competitor on their unified fabric vision.While there was plenty of talk about a common fabric integrating 8G FC, 10GbE+FCoE, over Common Enhanced Ethernet (CEE), the presentation and models focused on pure strengths of FC solutions between server and storage (maybe their ppt just haven’t caught up yet). At the end of the presentation there were many questions about this competitor’s 10GbE+FCoE capabilities, and how they see themselves fitting into a Cisco network already interconnecting the servers on the Ethernet side (…no, I didn’t ask any of those questions). So the presenter was feeling the pressure from the audience, and it was clear that the common interface was going to become Ethernet, and that fact was making them somewhat uncomfortable. So the vision is being presented by other vendors as well, however their ability to execute towards that vision is not clear. The presenter did promise additional product launches in the coming year that would enhance their CEE offering.Just thought I’d pass along…Regards,Marcin.

       0 likes

  2. It’s certainly something that I would like to explore. Beyond the reasons you state, of course, there’s another one that looms large over them all – the unwillingness to look at new solutions from new vendors. Traditional storage vendors as well as the competition (the B people, primarily) like to go on about Cisco being
    ew to the market”” and “”do you really want to trust your storage to those network guys?””In my organization, at least, I will have to overcome that kind of thinking long before the technical merits will have anything to do with the discussion.”

       0 likes

  3. Omar Sultan

    Marcin:Thanks for passing along your impressions. Over the last couple of weeks I have read in more than once that things are consolidating on Ethernet. The question now is the time frame–I think it will be longer than some expect. I agree with Doug’s assessment in a previous blog post: FC will still be around in a decade. With strong plays in Ethernet, FC, and Unified fabric, we can afford to be agnostic on this front.Actually, from a customer perspective, I think one of the good things is that pretty much everyone is working from the same playbook–listen to all the major players and you will hear common themes around consolidation, virtualization, automation, unified fabric, etc. This is a pleasant change from the past, where customers had to sort through multiple competing architectures and ran the risk of ending up in an evolutionary dead-end. Sure there is competitive differentiation, but companies are, for the most part, not moving orthogonal to each other.As you note, Marcin, a lot of it will come down to who makes the most effective transition from slideware to hardware.Regards,Omar

       0 likes

  4. Omar Sultan

    Steve:I certainly understand the challenges with this. I do think it is worth noting if a company starts offering FUD instead of better solutions.In most cases, it is simply a matter of education–at one time, we were new to the market, but that is certainly not the case any more in terms of market share or technology and thought leadership.Anyway, if I can be of help, please feel free to ping me directly.Regards,Omar

       0 likes