Tuesday, August 21, 2007

Stumping for Ethernet storage

Alacritech said Monday its product revenue from iSCSI sales doubled from the previous fiscal year, and the company expects it to double again in its next fiscal year. No doubt the actual revenue numbers are pretty low, since Alacritech declined to state them. You can assume this is one of those phenomena of huge growth on a tiny base.

The company also claimed sales are on the rise for Microsoft's TCP Chimney offload software. Alacritech has technology inside Chimney that Microsoft licensed after an intellectual property tussle. Again, no hard numbers. I welcome anyone with hard data to share it in a comment here or by dropping me a note at rbmerrit@cmp.com. From what I have heard anecdotally, use of Chimney is still very marginal because few end users are enabling TCP offload in their networks.

Alacritech quoted the Enterprise Strategy Group, a market watcher I don't know, saying more than 20,000 customers have deployed iSCSI. In a recent study, the analyst firm found that of 500 IT managers surveyed 17 percent are using iSCSI in product environments with 20 percent planning to implement the technology. In addition, Alacritech quoted International Data Corp as saying it expects iSCSI to account for 20 percent of external disk storage by 2010.

Lots of interesting factoids, but I take them with a grain of salt. Market research habitually shows everything moving up and to the right, and I suspect even in their optimistic reports there are more wrinkles and qualifiers to the story Alacritech chose not to quote.

Alacritech's "news" may be little more than marketing in an effort to attract a few more deals for its TOE and iSCSI technology. Still, the company should be allowed to crow a little bit. It helped pioneer those technologies and slogged through the desert of the past several years when it was not clear if either would take off.

Frankly we are not out of that desert yet. Alacritech may be seeing no more than a small oasis, or even a mirage in the distance. As I have written here many times, Intel and Sun are trying to obviate the need for TOEs with new processor features and Ethernet designs they are happy to license. And as for iSCSI, it has been the next big thing since about 2000. Many of the startups that raised millions on its promise are no longer here to pop a cork with Larry Boucher and Co.

If someday iSCSI and TOE do go mainstream, Alacritech and other survivors may be too busy really celebrating to issue a press release.


Anonymous said...

Not to quibble but:

TOE technology as pioneered or more accurately stated as many of the techniques used to create a TOE device, by multiple researchers in a variety of universities and companies over a very long span of years. There is no single company that pioneered the technology. It is true that this particular company was able to obtain a set of patents ahead of others that they have used to their advantage such as the licensing deal with Microsoft but that has more to do with what the Patent Office approves as patents than technololgy pioneer efforts.

Last I looked at the IETF iSCSI effort, the work was done by a large number of people from the storage companies and research with modest work from NIC providers. Having tracked the technology through out its development, I find it a stretch to assert that this company was a major player.

In terms of iSCSI's growth, it does have a market but the big question is whether FCoE rains on the tepid iSCSI adoption rate. iSCSI has many things going for it including the ability to run over any fabric or take advantage of new RDMA capabilities through the iSER wire protocol. In many ways it seems like the ideal protocol. Unfortunately, being versatile translates into being complex at various layers and that leads to slow product development and drag. For example, IBM just announced they are using Engenio's platform for an iSCSI solution while failing to highlight iSCSI has been shipping in targets from their competition for a number of years now. The press buys into this because the adoption rates for iSCSI are that low.

In any case, this particular company promoting what they purport as good news for their sales is all fine and good - more power to them. However, they are literally a drop in the bucket when it comes to the high volume NIC vendor solutions. What matters is what Intel and Broadcom do along with the new round of 10 GbE providers - ServerEngines, Chelsio, Neterion, Qlogic, etc. Those companies have considerable impact on bottom line sales and adoption rates. It is their response to the server and storage vendors that matters most in noting trends.

As for analysts, well, they get it wrong most of the time. One person I knew in marketing said the easiest way to get your company to take a particular action was to go find or pay an analyst to provide an analysis that said just what you wanted it to say. Woe be to executives who make decisions on such information but it seems to be a common practice. Just look at how much time people wrangle over such reports to insure the statistics are spun in their favor to understand the truth in such statements.

Peter Craft said...

As an Alacritech employee, I'd like to address some of the observations made by the anonymous poster.

Regarding the history of TOE, there's no question that research projects, and indeed products (remember Excelan?), existed long before Alacritech or any of the other current players in the TOE market.

What Alacritech introduced in 1997 was the notion of "Dynamic TCP Offload", or put more simply, the ability to migrate an existing TCP connection from the operating system to the network card and back. This enhancement eliminated numerous security and connectivity issues associated with "full-offload" solutions, and allowed for features such as link-aggregation to work.

As for the future of iSCSI - Note that all of the NIC vendors that you list, with the sole exception of Intel, claim to have or be developing TCP offload solutions. With FCoE years away at best, I'd say it's a safe bet that TOE will be pervasive by the time FCoE arrives. Once the overhead of TCP is offloaded to hardware, I fail to see what FCoE brings to the picture.

Anonymous said...

Well dynamic connection migration was definitely not new in 1997 though perhaps in a formal product it was considered novel at the time. People have being doing connection establishment and migration in NICs well before then both for TCP as well as a variety of other protocols though mostly focused on HPC applications.

What FCoE brings is several fold depending upon how much you drink the marketing hype kool-aid:

- Small fabric connectivity to legacy FC targets.
- SAN management tool extension
- Ability for FC vendors to ride the Ethernet performance train at the layer 2 / physical layer.
- Ability for customers to migrate without having to perturb their entire storage stack
- Some amount of cable consolidation for the rack server base.
Some amount of pricing flexibility for switches as FC switches have generally been higher priced than Ethernet
- Yet another opportunity for major switch providers to attempt to wrestle more control away from the initiator and target providers.

And so forth.

Peter Craft said...

re pre-1997 dynamic connection migration - I wasn't aware of that. Could you give me an example?

re FCoE - Most of the arguments you present argue in favor of FCoE over FC. You won't get any argument from me that, one way or another, ethernet is going to replace legacy FC.

What doesn't make sense to me is implementing another unreliable protocol on top of ethernet in order to overcome the overhead of TCP - particularly when the overhead of TCP is handled by TOE.

Consider other critical network applications - SMB, NFS, HTTP, all backup applications that I'm aware of - all run over TCP now, and with good reason. Reliable data deliver is best handled at the endpoints. FCoE may yet prove that three decades of conventional wisdom is wrong about this, but I doubt it.