In-Depth

An Open Letter To IBM

A note on the impact of iSCSI in the storage industry.

• By Jon William Toigo
• 12/02/2004

In this job, I tend to receive a lot of feedback from industry marketing folks who, after passing along obligatory kudos regarding some recent article or book I've written, ask me to schedule time for telephonic conference calls covering their latest products. I’m not complaining or anything, but my hard disk is getting pretty full of PowerPoint slides used to a facilitate these chats. Most are short on original ideas and long on “marketecture”, which strikes me as especially ironic given the all the vendor blather these days about “low data density” and “poor storage utilization efficiency” – basically the foundational “business case” that these very same vendors are using to drive “ILM.”

For what its worth, my recommendation for an effective vendor slide show is simple: get to the point fast. Skip the market research slides from Gartner, IDC, and the rest. I already know the numbers and can pretty well guess who paid for them. Just tell me the business problem you perceive, and give me a couple of diagrams showing how your product solves the problem. Then back-fill the presentation with a couple of charts showing cost, performance, manageability, and trade-offs that go along with the solution.

The last item is often the toughest for “marketeers", since it requires them to be honest. I want to know whether the product they’re pitching is a lock-in to the vendor’s technology, whether the same problem can be resolved in other ways, and whether there are any other real or potential “gotchas” in the solution. It is valuable to get this info up front, straight from the vendor, so that they won’t be surprised by the tone of the column that may result from their pitch. Moreover, by doing a bit of reflection, maybe vendors will finally see their products through the eyes of consumers and address any deficits in value proposition.

Now, in contrast to the above, every once in awhile, I receive an invitation to do something that I actually want to do. Case in point: a couple of weeks ago, Alain Azagury, a manager within the Storage and Systems Technologies practice at IBM Research Lab in Haifa, wrote to tell me that he was “in the process of building a package to recognize IBM's contributions to iSCSI in general, and Julian Satran's in particular as main author of the standard…for a self-assessment that IBM's Research Division conducts on a yearly basis.” Would I be willing, he asked, to write a letter highlighting the impact of iSCSI in the storage industry, as well as IBM's (and Julian's) contributions “to this revolution”?

Needless to say, I was honored and delighted to have the opportunity. I informed Azagury that I would not only write the note, but publish it as well, given the enormous esteem in which I hold Julian Satran. My letter was terse and is reprinted here in its entirety.

It is an honor and privilege to add my small voice to what is likely to be a stampeding herd of endorsements of the work performed by Julian Satran in bringing the iSCSI standard to fruition. I regret that I have personally met Julian only once and very briefly at the just concluded Spring Storage Networking World conference in Orlando, FL. I feel as though I have known him for years based on our many e-mail exchanges in which he provided both technical clarifications and “back channel” explanations about the politics of standards-making.

At a time when few vendors are willing to adopt a public vision of any sort on storage technology, it is great to have the ear of one of the most important technical resources for one of the largest vendors on the planet to address issues and questions in such a practical and down-to-earth manner. As a consumer advocate, I hesitate to endorse any vendor. That said, I think considerably more of IBM because of Julian.

Julian seems to eschew recognition, but he deserves it. If not for him, the iSCSI protocol would probably have foundered. Vendor obfuscation – led by the “Taliban” of the storage industry: vendor members of the Fibre Channel Industry Association – delayed the finalization of the protocol and taxed the patience of everyone involved in the process. With other early iSCSI evangelists falling by the wayside in opportunistic pursuit of revenues from cost-inefficient Fibre Channel fabric “solutions” (which are miscast as a “SAN” interconnect), only Julian and a few others stayed the course with iSCSI.

In the end, millions of small and medium sized businesses—and probably a lot of large enterprises too—will owe a considerable debt of gratitude to Julian and his tenacity. He is quite an asset to IBM, so don’t be surprised if this note ends up in one of my upcoming columns.

To this short missive, already delivered to Azagury, I would add one final thought. During my chance encounter with Julian at SNW, who is an unassuming gentleman of short stature and a 10,000-yard stare I’m guessing he acquired from years in the storage trenches, I complimented him directly on his work on the iSCSI standard at IETF. He just winked at me and said something to the effect that iSCSI was “easy.” Get ready, he said, for something coming shortly that will really be important.

I have no idea what he was referring to, but the remark confirmed for me that big developments are afoot in storage technology. As soon as I find out what Satran meant by his quizzical comment, I’ll share it with all of you. Until then, write to [email protected]