alphaspirit - Fotolia
Toigo Partners International
Published: 01 Aug 2014
Maybe it's a sign of the times: the earnings reports of leading storage array vendors are in the tank, product announcements tend to be humdrum "incremental releases," and unemployed smart guys are contacting me to ask if I know of any job opportunities. Makes you wonder if the Great Recession is really over.
A short time ago, an article in the San Jose Mercury News reported an uptick in plastic surgeries in Northern California. What distinguished this particular rise, the reporter noted, was that the folks in the doctors' waiting rooms are now predominantly men seeking procedures that will "de-age" them so they can find work among Silicon Valley startups. Turns out, if you're over 30 (and look it), it's assumed you don't have any fresh or innovative technology ideas left in your kit.
A lot of my friends are pushing 60 and can remember when it was their generation that declared anyone over 30 as untrustworthy -- so I guess it's a case of what goes around comes around. But a lot of these underemployed "old-timers" know exactly what the score is in storage technology and in IT generally. They have experience using technologies as well as experience developing new technologies, and they know how to evangelize and market technologies. It's a combination that affords a wise perspective about the way things are.
Perhaps what separates the old-timers from the newbies is a sense of realism. Having learned and re-learned the old saw about being able to "fool all of the people some of the time, and some of the people all of the time, but never all of the people all of the time," they question the industry's ability to boil the tech ocean with some half-baked notion of future computing. For every innovator who gets lucky, there are thousands of wannabes in the Valley who are there just to spend venture capitalist (VC) money. For every lionized startup CEO with a rags-to-riches success story, there are a lot of serial entrepreneurs who simply change T-shirt logos every couple of years and live off the kindness of greedy hedge fund managers in search of the Next Big Thing.
Technology, in the best of circumstances, evolves. Poor technology revolves -- like the insanity Einstein described of doing the same thing over and over again and expecting different results. Perhaps the lack of enthusiasm among older IT practitioners regarding software-defined storage, cloud and virtualization is that they recognize them as "revolutionary" in the actual sense of the word: Something we've seen over and over again in our careers, which is a recurring pattern supported by the VC community's desire for another tech bubble that will fill up their coffers until it bursts (leaving clueless IT executives and dead start-ups in its wake).
Here's an excerpt from an email I received from a friend of many years, who also happens to be a recently downsized corporate executive with a big software company.
Frankly I got off the virtualization believer list when I realized that we were creating massive complexity (managing in virtual vs. real space) to save money as server costs plummet. I concluded that the increase in management complexity needed to make a virtual world look real was not worth it. Most customers aren't running near as efficient[ly] as virtual promised, and the cost of virtualization software makes you wonder about sub-optimization vs. just buying more stuff or mov[ing] to the cloud. I am not convinced that most IT shops can afford to manage their stuff efficiently, that is why I think they will all be in the cloud in spite of the challenges of security, etc. I have yet to see an enterprise shop with elastic computing that worked or was worth the complexity.
The fellow makes me want to start up a new software company, one that eschews the current memes of the moment in technology and goes straight to the heart of the problem in IT: a lack of management of the infrastructure and the data that resides on it. Frankly, all the virtualization, cloud and software-defined stories I've been reading about lately strike me as half-baked efforts to solve the woes of utilization inefficiency -- in storage, servers and networks -- not by addressing the core issues of infrastructure and data management, but by using the plastic surgeon's techniques of virtualization to mask them from view.
If you listen to propaganda from VMware and even Microsoft around software-defined storage, you'd think de-evolving back to server-attached storage will magically solve the problems of storage scalability and provisioning. The wise old guys in the industry offer their "Yeah … buts" -- "Yeah, you can keep fielding direct-attached storage, but what's all that array-to-array replication going to do to your LAN bandwidth and end-user experience?" Or they might say, "Yeah, you can hide poor storage management by playing a shell game and replicating VMDKs and VHDs all over hell and half of Georgia, but what does that do to your storage budget besides grow capacity demand by 300% per year?"
It may sound like us old-timers are just throwing cold water on fresh ideas, but it's only because the ideas aren't really new or fresh. By and large, these "revolutionary" technologies reflect the soft bigotry of low expectations. VCs don't really expect startups to fix problems. They only care about spinning the innovation story enough to enable them to sell the startup at a huge profit to a big-name company.
I fear this is happening even in some of the smaller storage shops I've long admired for their obstinate dedication to fixing the problems customers know (but won't admit) they have and that the industry ignores because consumers demonstrate little interest in paying extra to solve them. Another email from an engineer friend, who is an expert at adding hardware-agnostic RESTful management APIs to storage products and was recently shown the door by his employer, expresses a sincere desire to do some good in the IT world "by extending RESTful Web services to connect the compute, network and storage domains together … laying the foundation for all of the cool capabilities" that the cloud folks can only talk about.
But early conversations with VCs have him worried the endeavor won't come to fruition because potential investors question whether a business model exists that will support the success of an infrastructure management play. It's so much easier to outsource problems to the cloud.
Meanwhile, companies like CommVault, SolarWinds, Tarmin and a few others slog on, often being forced to bend and twist an otherwise good narrative around data and services management into a kluge of marketing rhetoric that has enough money behind it to rate "search engine optimized" keyword status. From where we old-timers are standing, this has got to change or we will eventually have an IT apocalypse.
About the author:
Jon William Toigo is a 30-year IT veteran, CEO and managing principal of Toigo Partners International, and chairman of the Data Management Institute.