Home > Cisco, EMC, IBM, UCS, Uncategorized, VMware > FCoE end to end and the trough of disillusionment

FCoE end to end and the trough of disillusionment

Current customers with fibre channel environments looking to refresh their infrastructue always ask me one question, can we replace fibre channel with FCoE end to end for my whole environment.
For most customers, 100% virtualisation is a dream that wont be happening in the near future, replacing fibre channel with FCoE is in the same boat.
Unfortately unless you are replacing all of your compute, storage and connectivity end to end FCoE is simply not possible and if you have non-virtualised workloads its even further away!
I have vendors consistently telling me they have customers going 100% FCoE only to find it a small environment where the requirements are specific enough to make FCoE end to end happen.
Unfortunately we are at the stage where hype/vendors/expectations dont meet expectations and that lands us fair and square in the trough of disillusionment.

Categories: Cisco, EMC, IBM, UCS, Uncategorized, VMware
  1. December 24, 2011 at 11:56 pm

    I agree that the adoption of FCoE has been slower than was originally anticipated. As to your other points:

    1. I don’t understand why you are tying virtualization and FCoE together? Non-virtualized hosts work just as well with and without FCoE. Can you explain?

    2. Customers do not have to replace compute and storage. If you are using a fairly current platform, upgrading to FCoE can be as simple as swapping out an FC adapter for one that supports FCoE.

    3. The entire network doesn’t need to be replaced. However, you’ll need to attach the host CNA (FCoE HBA) to specialized FCoE switches (FCFs). Today the CNA will most likely be attached directly to the FCF but as more solutions come into the market, there will be additional choices. This may be the biggest stumbling block for people and I can understand why this is viewed as a limitation.

    4. I’m aware of companies moving in the direction of complete FCoE end to end but nobody (that I’m aware of) has gone all the way yet. This doesn’t mean that FCoE isn’t ready, it means that they don’t see the need to move existing workloads off of platforms / networks that are working perfectly fine now. (If it ain’t broke, don’t fix it)

    5. I think it’s possible that FCoE has entered into the “trough of disillusionment” and if this is the case I wouldn’t be surprised to see it ascending the “slope of enlightenment” in 2012 once people start to appreciate what’s possible once you start migrating to it.

    Disclaimer: I work for EMC and have been working on FCoE for over 4 years both as a system integrator and as a participant in the FC-BB-5 and FC-BB-6 working groups (responsible for the FCoE Standard)..

    • January 1, 2012 at 1:28 pm

      Hi Erik,

      Thank you for your input and comments, I appreciate your point of view. The following are my comments to each of the points you raised.

      1. Linking virtualisation and easy adoption of end to end is primarily due to the reduced cost in connecting fewer hosts via CNAs to the fabric, only recent compute hardware purchased in the last twelve months come standard with FCoE capable adapters. Most customers I deal with who have high ratio of virtualisation also have high consolidation ratios and therefore much fewer hosts.
      Also the impact and changes for customers with higher virtualisation rates are lower as they need to do less compatibility work with drivers and operating systems than customers with more physical hosts. Having a standard hypervisor in your environment simplifies the changes and risks as only one hypervisor and the CNAs drivers need to be tested by abstracting away the storage change to the VMs.

      2. A large majority of my customers use blade chassis and therefore although some of the their new blades will have onboard adapters capable of FCoE their platform with require significant upgrades to the chassis connectivity. Customers with rack mount systems may have an easier upgrade path but it comes back to the number of hosts and extent of FCoE capable infrastructure they already own.

      3. Every customer who asks me about moving to FCoE use the retirement of either their current FC fabric and/or their storage as a trigger to make the switch. In most cases the complete retirement of a FC fabric in favour of a unified fabric requires compute connectivity changes and network upgrades that have no been budgeted for by the customer. When its more practical to maintain a smaller FC fabric for legacy hosts most customers seem reluctant to have three networks (Ethernet, FC and FCoE).

      4. I agree the FCoE for host connectivity into the network is a great option for customers and this is generally what I recommend. There are a few small examples of end to end FCoE but they are mostly vendor sponsored to prove it can be done. I’m not claiming FCoE is not ready for host connectivity, just FCoE end to end for most customer environments.

      5. I’m also hoping that more customers move towards FCoE but I believe most FCoE with be for host connectivity and not a complete end to end architecture.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: