Home | Featured | Enterprise Computing: Is 2010 The Year for iSCSI or FCoE? Place Your Bets!

Enterprise Computing: Is 2010 The Year for iSCSI or FCoE? Place Your Bets!

0 Flares Twitter 0 Facebook 0 Google+ 0 StumbleUpon 0 Buffer 0 LinkedIn 0 0 Flares ×

It’s pretty traditional to start the year looking at what will be prevalent in the coming 12 months.  In this opinion article, I will discuss whether we’ll see the prevalence of iSCSI or FCoE and what we can expect, looking back this time in 2011.

History and Background on iSCSI

iSCSI as a concept was first developed by IBM in 1998.  With work from Cisco, a first version of the iSCSI standard was presented to the IETF in 2000.  Today we see iSCSI implementations in enterprise storage arrays from the major vendors, from mid-range storage array manufacturers and in virtual appliances.  The iSCSI initiator (the host component for making connections to an iSCSI device) is standard in major operating systems such as Windows and Linux.  In fact, some vendors like HP/Lefthand and Dell/Equallogic make a virtue of deploying their technology as iSCSI only solutions.

However despite the ubiquitous nature of the iSCSI technology, I don’t think the adoption level is particularly prevalent in the majority of enterprise and mid-range environments, with fibre channel remaining the dominant player.

History and Background on FCoE

By comparison, Fibre Channel over Ethernet is a relatively new technology.  In fact, development to ratification has only taken two years and was approved in June 2009.  FCoE enables the fibre channel protocol to be passed over 10Gb Ethernet natively.  This is a distinct difference from iSCSI which relies on using the TCP/IP stack for communication.  As FCoE is so new, it isn’t widely adopted or supported by vendors today, with notably Cisco, EMC and Netapp being the main leaders in offering products or announcing future support.

Why iSCSI Has Failed

It’s a bold suggestion, but I personally think iSCSI has failed to live up to the hype surrounding its development.  Although iSCSI is cheaper to implement and support, Fibre Channel remains the dominant force in storage networking today.  Why is that?  It’s a question I’ve answered many times, there are a number of reasons:

  • Legacy of deployed technology. Fibre Channel became entrenched in storage architecture before iSCSI arrived on the scene.  Many organisations made investments in skills and hardware, which once established made FC the de facto protocol for storage networking.
  • Hardware interoperability. iSCSI and Fibre Channel don’t inherently work together.  As far as I am aware, Cisco were the only major vendor to offer a product that had any kind of FC/iSCSI bridge functionality.  This makes the two protocols pretty distinct and there’s no benefit to running multiple disparate storage protocols within the same environment.
  • The Cultural Issue. Fibre Channel devices have typically been managed by the storage team.  iSCSI and IP are the domain of the networking team and “Never the Twain Shall Meet”.  Even though iSCSI is recommended to have a dedicated infrastructure (which should still work out cheaper than FC), the hardware configuration and deployment would be managed by the network teams.  There are very few instances I’ve seen (and certainly not viable ones) where Storage and Networks work in harmony.

Why FCoE will Succeed

Where iSCSI has failed, FCoE will surely succeed.  As a protocol, it offers more opportunity of integration between “traditional” fibre channel and FCoE.  In fact, hybrid devices are already available from the market leaders, Cisco and Brocade.  FCoE is being built into the hardware designs of integrated solutions from Cisco and Acadia.  Solutions from HP and IBM will have to offer the same functionality.  The whole culture issue will be removed by the deployment of end-to-end architectures.  Perhaps in addition the two teams may just have to learn to live with each other.

A Mixed Landscape

I think in the next 12 months we’ll see the gradual adoption of FCoE as legacy FC environments come up for renewal and the choice between 10Gb/e and 10G FC has to be made.  There will no doubt be a lot of confusion as to which protocol suits large environments best.  My negativity towards FCoE has been based on exactly what benefit FCoE would provide in large environments over FC.  After all, cost is hardly the issue when FCoE CNAs are still so expensive.  Overall, iSCSI will be the loser and will be relegated to mid-range and SOHO solutions where it does an excellent job.  In 12 months time, we’ll be discussing the battle between FCoE and FC and iSCSI will be an also ran.

About Chris M Evans

Chris M Evans has worked in the technology industry since 1987, starting as a systems programmer on the IBM mainframe platform, while retaining an interest in storage. After working abroad, he co-founded an Internet-based music distribution company during the .com era, returning to consultancy in the new millennium. In 2009 Chris co-founded Langton Blue Ltd (www.langtonblue.com), a boutique consultancy firm focused on delivering business benefit through efficient technology deployments. Chris writes a popular blog at http://blog.architecting.it, attends many conferences and invitation-only events and can be found providing regular industry contributions through Twitter (@chrismevans) and other social media outlets.
  • http://blog.scottlowe.org Scott Lowe

    Well, I guess I have a 50/50 chance of guessing right, so I’ll put my prediction behind FCoE. In addition to any technical differences, FCoE just seems to have too much attention, too much vendor support (read: Cisco) in order to not make some significant inroads this year.

    So, there’s my guess–we’ll see next year if I was right or wrong!

  • http://www.chriscowley.me.uk Chris Cowley

    My opinion is that it will be FCoE, although I am not certain given the current economic climate that it will be this year.

    I am in the situation that I have a 5 year old 2Gb FC SAN. This needs replacing, but the money being there is a different matter. What is almost certain is that we will stay with FC. Hence we will probably have a certain amount of FCoE.

    My suspicion is that we are not on our own, thus my weight woul dbe behind FCoE, but we are jumping the gun by predicting it to be this year.

  • http://www.chriscowley.me.uk Chris Cowley

    Also, with respect to your comment that network and storage teams never meet – I disagree.

    I work in a pretty small team of 5 (including our boss) 2 of us look after Linux and infrastructure. Infrastructure covers LAN, WAN and SAN.

    See – storage and networking people can not only talk to each other, but can even be the same people.

  • http://www.brookend.com Chris Evans

    Chris
    Thanks for the comments. One thing I didn’t mention was price. A quick check on street prices for Emulex LP21000 CNAs to LP11000 HBAs shows the CNA cards at around £1200 and HBAs £600 (obviously resellers and bulk buys may be cheaper).

    This differential needs to change – unless of course you run a mixed environment and retain CNAs purely for consolidated servers.

    Oh and I get your point on small teams working together; that’s absolutely true. Imagine wearing two hats and arguing with yourself all day!

    Chris

  • Paul P

    I’m not convinced one has to beat the other, or that they are mutually exclusive. Just as we have iSCSI and FC today.

    What I predict happening is FCoE replacing FC in the big end of town and iSCSI continue growing in many market segments, including the small and medium (the volume markets) – at least until 10Gb ethernet is standard in even small offices (which is surely still a while away).

    I would have thought FC is most at risk here not iSCSI.

  • http://blogstu.wordpress.com Stuart Miniman

    Chris,
    Where I fully agree with you is that “hype” tends to drive perceptions too much in this space. iSCSI was originally pitches as a “FC-killer” and it has not done this. For years I spoke with the large enterprise customers in EMC’s customer base and iSCSI was not something that interested the storage administrators. That being said, last year at my FCoE session, in an audience that had 100% FC adoption, there was also 1/3 of the audience that had iSCSI deployed. Yes, iSCSI is today in smaller shops (where the administrator handles the servers, network and storage) or in small pockets of the larger enterprise. FCoE fits nicely into the management paradigm of FC and therefore can move into those big configurations much easier than iSCSI. The real challenge is moving from the standalone SAN (either FC or iSCSI) into a converged environment where LAN and SAN are on a single wire. We need this for cost; while CNAs are not inexpensive, the overall Capex and Opex of the environment can be improved by moving to an environment where 100% of servers have the option for SAN attach rather than the current model where we choose whether to use network storage on a server by server basis which doesn’t allow for the flexible, virtualized environment that customers are moving to. Let’s not over-hype what FCoE can do TODAY – it will take a few years to work out the technical and political issues. There is also plenty of room in the marketplace for iSCSI and FCoE to flourish.

  • http://www.dell.com Robert Winter

    If iSCSI has failed it has been a magnificent failure with increased shipments every year (refer to your latest industry analyst charts – any of them). The arrival of iSCSI offload and 10 Gbps Ethernet has given iSCSI the needed boost to outperform FC. When iSCSI was a 1Gbps Ethernet technology FC had an advantage with 4 Gbps media. The reason FC has gone to FCoE now is that the speed of Ethernet has outstripped FC and will continue to do so with future 40/100/400 Gbps deployments. But, and hear’s a big but, Ethernet isn’t FC and you can’t just ignore the quality of the FC physical and data link layers (FC-0/1) that provided a superior (in my opinion) flow control mechanism to Ethernet by using proactive credits and not the tried and not always true Ethernet method of reactive PAUSE frames. The new DCB (Data Center Bridging) Ethernet standards from IEEE, in particular the PFC – 802.1Qbb standard, just replicate this PAUSE mechanism 8 times, once for each priority on a link. Not sure that really helps anything. iSCSI, however, can recover from a dropped frame – it has a 25ms TCP fast-retransmit capability. FCoE doesn’t have anything like this and will incur an I/O (60 second) timeout if ANY frame is dropped. So, with FCoE you have the option of dropping frames or building in over-capacity in your networks so there’s no congestion. Also the relative (to iSCSI) immaturity of FCoE standards continues to be a concern as maturity only comes with deployment history. iSCSI has been deployed for almost a decade and is improving by continued and accelerating use. The one deficiency of iSCSI relates to its integration into Enterprise class management systems but that is due more to the fact that Ethernet itself is not fully integrated into a storage mangement context – and FCoE, by nature, has the same problem. Maybe FCoE will be there in a few years – but it’s not ready for prime time yet – despite all the hype.

  • http://www.chriscowley.me.uk Chris Cowley

    “Imagine wearing two hats and arguing with yourself all day!”

    Seeing as I work in Portsmouth and my main colleague works in Lincolnshire then I sometimes have to resort to that just to have a good argument!

  • http://nigelpoulton.com Nigel Poulton

    Chris,

    It will be interesting to see how iSCSI benefits from 10Gbps Ethernet. Some of the DCB extensions, in particular Priority based Flow Control (PFC), can extend lossless functionality to iSCSI – anybody for iSCSI over UDP rather than TCP. UDP might be viable on a lossless Ethernet!?

    CNAs are also starting to offer hardware offloads for iSCSI too. Couple that with 10Gbps and performance *might* come up to similar levels.

    While I’m sure FCoE will pick up pace in 2010, I wonder if 2011 might be the year of FCoE.

    I dont doubt the benefits FCoE promises (consolidation, fewer cables, fewer I/O adapters, switch ports…) but the relative immaturity of the protocol and the economic client might hamper deployment this year. I think a lot of people will wait until the economy recovers a little more and the protocol and shipping products mature some.

    As for the comment saying FCoE might kill FC. Not sure what that means…. its still FC protocol encapsulated in a jumbo Ethernet frame. Management most of the protocol specifics stay.. in fact the only thing that springs to mind that will disappear is buffer credits – replaced by PFC.

    Nigel

  • Pingback: Tweets that mention The Storage Architect » Blog Archive » Enterprise Computing: Is 2010 The Year for iSCSI or FCoE? Place Your Bets! -- Topsy.com()

  • Joe Galietto

    My opinion is that we will see iSCSI with 10GB Ethernet grow in the Small and Medium business segments. It will grow in the shops with the smallest IT staffs. I do not think we will see large scale commitments to FCoE in the first 3 quarters of this year. Firms which are using FC today and who need to refresh this year I expect to make a non-decision and go with 8GBs FC. That may change in the 4th quarter if FCoE matures sufficiently. If FCoE does not gain significant traction by Q4 it is effectively dead. But, that does not mean iSCSI is the winner in the enterprise.

    I think that FCoE is going to be an extremely tough sale because of internal politics at the customers storage vs network. The chorally of that is FCoE deployments are going to be beset with failures caused not by the technology by political divides at the customer.

    Joe

  • http://www.deepstorage.net Howard Marks

    You’ve basically got why iSCSI didn’t succeed in the enterprise right. Two solutions to the same problem is more expensive, even if the incremental cost of the new tech is lower and turf. I think you’re underestimating the turf battles around who owns the FCoE switch.

    iSCSI has a longer life ahead than native FC. Organizations that now have 1-6 FC switches (as opposed to directors) and no real SAN management tools (SANscreen Etc.) are just as likely to replace that infrastructure they never really understood with 10gig iSCSI as FCoE. FCoE is a good solution for folks with hundreds of FC ports as they can keep the management processes and tools they have.

    A Nexus 5000 or Brocade 8000 costs 4x what a ProCurve or Foundry 10gig switch w/o FCoE does on a per port basis. That’s going to matter to some especially in a downturn.

    All should note that both FCoE and iSCSI (10gig) have been demoed to deliver a million IOPS so performance is an issue for only a small number of users. iSCSI performance has always been better than perception.

    – Howard

  • Pingback: chriscowley » Enterprise Computing: Is 2010 The Year for iSCSI or FCoE? Place Your Bets!()

  • Pingback: 10Gbit 以太网 FCoE 与 iSCSI 之对决 (FCoE vs. iSCSI. vs ATAoE) : 弯曲评论()

0 Flares Twitter 0 Facebook 0 Google+ 0 StumbleUpon 0 Buffer 0 LinkedIn 0 0 Flares ×