Toigo’s Tomatoes 2005
If you don't have something nice to say, say nothing at all, right? Storage columnist Jon Toigo ignores that advice as he looks at the three worst events and trends in the storage industry this year.
In last week’s column, I awarded my picks for best storage industry plays of the past year. My editor suggested that I consider doing a column on the worst storage-related items and issues we’ve seen over the past 12 months, “Toigo’s Tomatoes," so to speak. Our parents always told us that if we had nothing nice to say about someone, say nothing at all. Clearly, our parents didn’t know the storage industry. The challenge of selecting Toigo’s Tomatoes for 2005 is winnowing down the overwhelming list of valid candidates for the award.
Generally, I prefer not to single out specific companies for Tomatoes, but rather obnoxious trends in technology or "marketecture" that have earned my ire. Let's get started.
Just Say No
Sadly, 2005 saw the introduction of yet another instance of a consumer gag order imposed by a storage vendor when Network Appliance added a codicil to its Filer software license prohibiting customers from speaking publicly about the performance they see from using their product. Basically, if you go public with your performance issues, you can no longer legally attach your NetApp Filer to a network because you are no longer legally within your rights to run the software required to access the box via the Network File System (NFS) or Microsoft SMB across your LAN.
EMC has similar provisions on its high-end gear. If you speak publicly about performance, you negate your warranty.
This practice must stop if consumers are to have any hope of making rational decisions about the storage products they buy. As a VP of storage at a major bank once complained to me, “Where are we supposed to get good information to guide our product selection? You know the vendors are lying because you can see their lips move. The analysts are in the hip pockets of the vendors—the pocket that holds the wallet. The trade press doesn’t have the resources to do a deep dive into products, and they need to worry about ad revenues. And the last round of IT budget cuts eliminated the test labs.”
If you can’t ask your neighbor about his experience with a product—or more precisely, if he can’t answer your questions because of a vendor gag order—from where, exactly, is that kind of practical, in-the-trenches information supposed to be gathered?
When I asked NetApp about the rationale for this practice, they told me that it was intended to prevent people from making bad reports on Filers that result from the way the consumers have deployed them or are misusing them. Mustn’t have those stupid storage managers dissing products when the poor performance they are receiving is their own fault!
NetApp, I was told, tests its gear and publishes the results. In fact, they test every Filer under SPEC.org’s SPECsfs regime (you can go there to see the results—http://www.spec.org/). They deftly used the opportunity to criticize EMC for not subjecting its gear to any independent testing.
I checked SPEC.org, and sure enough, the findings were there. Interestingly, however, their numbers didn’t match their current marketing claim that each new generation of their product doubled the price/performance value of the previous generation. I called them on that point and they explained that I didn’t understand how they calculate price/performance value. They promised to send me the calculations they were using—but the information never showed up.
I was running a “NetApp Data Held Hostage Day ##” thread on my personal blog at DrunkenData.com ( http://DrunkenData.com for several weeks, an amusing vigil that kept everyone in stitches for awhile until it got boring and I dropped it. Then, in October, at Storage Networking World, I ran into one of the founders of NetApp and asked about the price/performance claim. His response: “We said that? Well, it’s just not true. We’ve been hiring a lot of new people lately. Some of them are pretty strange.”
Bottom line: A rotten Toigo’s Tomato for 2005 flies out to Network Appliance and other vendors who use lame gag orders in their sales agreements to stop the free exchange of practical information about storage products. My advice to readers: refuse to sign a contract for any storage product that has such provisions unless the company draws a big fat red line through its provision and a company official initials it, letting you off the hook. Most of them will do it—they just want the sale.
Vendors Posing as Consumers
The second Toigo Tomato goes to vendors who misrepresent themselves as consumers of a competitor’s gear in order to bad mouth it in message boards, blogs, and other resources consulted by storage decision-makers. I had my first taste of this practice on an EMC stock-trading message board a couple of years ago, when Blue Arc was making its claims about being faster than EMC, NetApp, and every other product in the market.
Guys from Blue Arc were actually pretending to be EMC customers and letting loose a tirade of complaints about the competitor’s product performance “in our shop.” I did the trackbacks on their IP addresses to confirm that they were Blue Arc employees, strictly in the interests of fairness and not out of any special concern for protecting EMC’s less-than-stellar reputation or product performance; I confirmed that the comments were coming from Blue Arc and I called them on it.
The person to whom I spoke said, with a chuckle, that he was aware that some messages had been posted by some overly enthusiastic employees. (Those kids!) They had been told not to do that anymore. He didn’t seem to perceive the legal ramifications of the behavior—or maybe he was subtly acknowledging something that I have recently begun to realize: a lot of storage companies do the same thing. Sort of an underground smear campaign often employed to supplement above-board marketing initiatives in which direct attacks on competitor products are considered déclassé, gauche, or downright vulgar.
The real damage of this activity is to the consumer who is looking for valid information on which to base product selections. If you can’t trust that a fellow user is who he says he is, you can’t trust anyone. History tells us that tyranny thrives in such an environment, and propaganda is substituted for the free exchange of ideas and opinions based on empirical evidence.
Piles of Propaganda
Speaking of propaganda, the final Toigo Tomato of 2005 goes to the Storage Networking Industry Association, purveyor of the biggest piles of propaganda I have seen in years. Their just released End User Council Survey at http://advisorygroups.snia.org/home/SNIA_EUC_2005_Survey_11-09-2005-fnl.pdf is a fascinating read—a case study in propaganda.
Not only is the survey not a statistically relevant sampling of end-user opinion, the interpretation of the responses received is so stilted as to be laughable.
Of the 252 responses (a large percentage of which were from SNIA members or affiliated with SNIA), 54 percent were neutral or negative toward the statement that SMI-S would make any difference in storage interoperability problems. That's not much of an endorsement of the group’s much touted Storage Management Initiative Specification. Of the 46 percent who believed SMI-S would make any sort of difference, I noted the remark in the document that “the members of this group were more likely than not to belong to the SNIA End User Council.” In short, they were towing the line regardless of the deficits of SMI-S.
On SNIA’s Information Lifecycle Management initiative, which deserves a special Tomato of its very own, 61 percent responded "neutral to strongly disagree" regarding the SNIA definition of ILM. This should have given the organization a clue that their ILM “roadmap” was being correctly perceived as serving vendors.The SNIA ILM roadmap begins with solving all of the interoperability and low-level management problems in storage, then—many, many years in the future—provides an examination of business processes or definition of data classes. This lets vendors sell more hardware stovepipes rather than attacking the core problem ILM seeks to solve: unmanaged data growth. In a nutshell, the roadmap is upside down.
Rather than taking a clue from its own numbers, SNIA’s interpretation of the results is purely self-serving: “This question showed a high rate of neutral responses, which may indicate a lack of broad user understanding of this issue.” In short, users don’t see things SNIA’s way because they aren’t properly educated.
So, in 2005, SNIA created its own re-education camp in the storage gulag: the SNIA Training Continuum. The Continuum is endeavoring to present itself as the only valid purveyor of certifiable education on storage networking. I’d toss a tomato to these guys, too, but I won't waste good food on such nonsense.
There you go. Write back, with or without tomatoes, to firstname.lastname@example.org.