FCoE or not FCoE?.. that is the question

So I spent a day at Cisco last Friday for a cram training session on Cisco’s new Nexus range of switches, aside from some very impressive features of the switches by way of flexibility, speeds and feeds (12 Terrabits per seconds capability of you’re interested) and some very cool virtualisation features in the switches; one thing that came up was FCoE (Fibre Channel over Ethernet) which effectively encapsulates a Fibre Channel frame into an Ethernet frame with no extra frills,  no additional clunky IP protocols..  keep it simple, keep it quick, as opposed to iSCSI which is used to carry SCSI commands and provide block level storage connectivity over TCP/IP which is prone to packet loss. FCoE uses something coined enhanced Ethernet and something called priority flow control where packets are embedded with Class of Service (CoS) value.. where packets with a certain CoS value (namely FCoE packets) utilize no drop, reliable behavior and other packets use the familiar ‘Best Effort’ Policy as traditional Ethernet does. This means that Ethernet can be a viable medium to carry fibre channel traffic within the data center, where iSCSI is not.  Don’t get me wrong, iSCSI still has its place, FCoE is currently only supported over 10GbE and it will be a while until many organizations adopt 10 GbE fully.

Currently FCoE is geared more to the datacenter, the idea being that less cabling is required within the datacenter rack as both storage connectivity and standard IP traffic can be delivered through one form of connectivity; no need for NICs and FC HBA’s…  Converged Network Adapters or CNA’s are the new flavour (basically a NIC which supports FCoE). This means that server build times are reduced, doubling up on switching infrastructure to support FC and IP traffic is no more, cabling is reduced and everything is generally tidier.

There are still arguements against FCoE..   to site one example from the blog Etherealmind.com and the flip coin of the arguement from a Cisco Chappy Dante Malagrino (see here for his response to the previous link), so its not to the taste of all…

Have a dig into it and make your own judgements.

Some good sites of reference :

http://www.fcoe.com/

http://features.techworld.com/storage/3306/fcoe–or-iscsi/

Advertisements

About interestingevan

I work as a Technical Architect for a Storage and Virtualisation distributor in the UK called Magirus. The goal of this blog is simply to be a resource for people the want to learn about or go and Sell storage. I'm a qualified EMC Clariion Technical architect, Commvault Engineer and Cisco Unified computing specialist. I have also worked with the rest of the EMC portfolio for a good few years. This Blog will provide information on how specific technologies work, what questions need to be asked in order to spec certain products, competative info and my two pence on some of these technologies. Please feel free to provide feedback as to the content on this blog and some bits you'd like to see. View all posts by interestingevan

2 responses to “FCoE or not FCoE?.. that is the question

  • Ryan B

    Been wrestling with this question for a new build-out and I’m leaning towards iSCSI. I view FCoE as a transitional technology for consolidating ports at the edge while leaving an FC core in place. From a technical perspective the real magic is in Data Center Ethernet – if you can make iSCSI lossless by running it over DCE, why use FCoE?

    • interestingevan

      You’re right, currently FCoE sits very well at the access layer and will enable a business with a significant investment in an FC SAN fabric to consolidate IO at the edge, but still make good use of existing skillsets they may have inhouse around FC. In terms of an end to end solution, yes the capability is there. EMC have FCoE modules in the CX4, CNA cards now have FIP, the Cisco nexus 5000 has been around for a little while… but admittedly i might be tempted to wait until the vendor propositions have matured somewhat and FCoE is slightly less proprietory, especially as 802.1qbb hasn’t been officially ratified as a standard as yet.

      I much like you am still getting my head around all the applications of FCoE. One of the main benefits seems to be the payload size within the ethernet frame being greater than that of iSCSI as iSCSI still has a fair chunk dedicated to TCP/IP, so effectively you get more bang for your buck in terms of throughput (although that also seems to be a point of contention). My view is that FCoE as an end to end solution still has a little way to go but for IO consolidation at the edge of a DC which already has an FC SAN fabric.. very viable and a good shout.

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: