Storage management software is hot. Over the last year the industry has witnessed significant advances in storage management software technology, and an A-list of hardware companies from EMC Corp. to Sun Microsystems Inc. and Compaq Computer Corp. continue to scramble to add new storage software tools to their separate arsenals as hardware margins plummet. In the midst of this shift from emphasis on storage hardware to storage software, Veritas Software Corp.’s President and CEO Gary Bloom can smile, content with the fact that the Mountain View, Calif.-based software company has never manufactured a piece of hardware in the company’s history. InfoWorld senior writer Dan Neel caught up with Bloom at the recent Storage Management 2002 trade show in Chicago, to chat about just how it feels to have been a pure-play storage software company all along.
InfoWorld: Veritas has been a pure-play storage software company from the start. So now that storage hardware companies like EMC are rushing to re-make themselves as software vendors, how does it feel to already be where the action is?
Bloom: Well we think it’s great. In a funny way it’s nice of the big hardware vendors to come and validate what we’ve always believed, which is that all of the value is in software. And so having EMC and Sun and everybody come in — they’re all kind of saying “Yeah, software is the future of storage” and I fundamentally agree with them. I think we’re in an extremely unique and favorable position given that we have a broad range of storage software assets, from the operating system level of components up to all the new SAN technology and SAN virtualization technology. I also think that realistically, the barrier to entry for them to quickly become a software company is very large. First they have to evolve from the hardware company to be in software to begin with, and then they have to have their engineers build software that works across a broad range of network architectures, data base architectures, storage architectures. It’s tough.
InfoWorld: Do customers see an advantage by going with a software-only company for their storage management software?
Bloom: I don’t know if the customer really thinks about whether they prefer hardware companies to software companies. What they prefer is a solution that works in multiple hardware environments. And interestingly, if you look at that as the requirement, nobody else does that other than Veritas.
InfoWorld: Do companies in general automatically buy their storage products from their server vendor?
Bloom: I don’t think there’s a reality right now to the idea that says if you buy a certain server you automatically buy some storage; if you buy an IBM server you automatically buy IBM storage. The market data doesn’t support that, and it doesn’t because EMC, over so many years, simply got too big. And EMC became a very large vendor of storage, yet they never produced a server.
InfoWorld: Clear up the blurred definition of storage virtualization for us.
Bloom: Well, my belief is that at the end of the day virtualization is nothing more than what we’ve been doing for years, and it’s called volume management technology and you know it’s real simple. The only thing that’s changed, the only thing new in virtualization if you want to look for something new as a new term and a new issue, is it’s moving into the fabric because storage is moving into the network. So the conversions of storage and networks is simply moving the virtualization from the server called volume management into the network. But, fundamentally, the concept itself is not new by any means.
InfoWorld: What are storage customers asking for right now as far as added features to their management software?
Bloom: I would say there [are] probably three major agendas that customers have, and let me name them and then I’ll come back to each of them. One is the protection of data and the whole disaster recovery category. Another is reducing the amount of head count required to maintain and support their storage environments and their overall IT environments. And the third one is help us get better utilization of storage assets that we already own. Now let me back up on each of these. We really do think it is a disaster recovery market, it’s not just backup recovery. Backup recovery was a paradigm historically used to back up data and protect it from disasters. That was the paradigm and it said you put it on a tape and you send it offsite and that’s backup and recovery and it was designed for the idea that you could accidentally lose a storage device or, in the worst case, a water pipe breaks in your data center and you have a flood or you have a fire in your data center. Sept. 11 changed the whole paradigm of planning, and when that paradigm was finally changed, I believe disaster recovery planning also changed, going from more traditional, what I call offline backup, [where] we put on a tape and ship it away, to on-line backup, which is a fall-back site. & [Customers] can leverage the lessons we learned post [Sept. 11], having worked with so many customers, and then they can also, in many cases, take advantage of equipment and storage devices and servers that they may have already in their shop. They don’t always have to go in and think of it as $50 million of capital acquisition to buy another whole storage farm.
As for the cost to add additional storage and add additional storage technology, whether it be tape or disk, you know, I think there’s some industry estimates out there that it’s $4 to $7 in people cost for every dollar spent on the hardware, and that’s an outrageous cost. & Let’s get the people and labor costs out of that environment. Let’s make it cheaper, make it easier to manage [so] you don’t have to have an army of people to back up an army.
And then the last one has a lot to do with the storage utilization rates, which the Gartner Group estimates that storage utilization rates are no higher than about 50 percent. Most of the CIOs I ask & come in at 25 to 50 percent. Most well under 5 percent.
InfoWorld: Why so much excess capacity?
Bloom: Chances are if you’re buying 10TB, you really only need for maybe 7TB or 8TB. You buy some spare capacity and then when your favorite storage vendor walks in and says “Well, if you close the deal this quarter I’ll give you 14TB for the same price as 10TB.” Well, you say, that’s a no-brainer. And remember, I only needed 8TB to begin with and now I’ve got 14TB. I haven’t talked to a customer in six months that believes that they will consume all of their capacity.
InfoWorld: What happened to the anticipated global explosion of data?
Bloom: Administrators see the growth of data. The data volumes are increasing, meaning we’re collecting more — your day-to-day e-mail, this interview, everything eventually gets digitized [and] stored somewhere, and that’s creating massive hard data. But all this massive data volume growth is predominantly consuming excess capacity and potentially it will consume excess capacity for some time. And that’s why the hardware vendors, I think, are not going to see the full benefit of an economic recovery. I think the hardware vendors are going to be challenged because you have customers [whose] thought process about forward buying has changed, anyhow, on the hardware market and the software market.
InfoWorld: Has the urgency to deploy a disaster recovery cooled any since Sept. 11?
Bloom: I think people are always looking for this immediate skyrocket of technology purchases, kind of like every time there’s a virus that goes through, it’s the end of the world, everybody runs out and buys their favorite anti-virus software-insulated update and there’s this immediate up-tick in volume. Backup and recoveries are a much more complex equation to solve, and I think all we’re seeing is that people are now starting to understand the complexity of it and they’re stepping back from the emotional response to [Sept. 11]. They’re saying “I actually have to architect and design a strategy for disaster recovery for the future.”
InfoWorld: What will an economy of Web services eventually require from storage software?
Bloom: With Web services, you’re kind of taking what the dot-com world was supposed to be and what it’s actually becoming, which is Web services, and the only difference — the predominant difference — is it’s just simply becoming much more traditional companies taking part. The era of the dot-com is still happening, it’s just now you know it’s not going to be WebVan, it’s Safeway and Albertsons and all the local grocers that are starting on-line ordering and on-line deliveries, and in many cases they’re not experts in processing payments that way so they often do go to a payment service, some Web service. I think the intersection for Veritas and Web services is around business continuance and high availability. You process payments with a third party and in that scenario the key requirements of those Web services are high availability, and so when we talk about high availability that suggests very highly available storage architectures, sent around to both replicated sites and, more importantly, cluster sites. To me those are the ultimate levels of business continuance and at the very top of the business continuance curve. There can be no disruption of service, it’s continual availability, and those kinds of providers and those Web services will absolutely require those kinds of storage management software services.