STORAGE AT MIDYEAR.
One of the highlights of each year's Network Storage Conference is the Executive Roundtable, moderated by the Conference organizer--Farid Neema of Peripheral Concepts. This year's panel included Dave Weiss of StorageTek, Mark Leslie of Veritas, Jeff Allen of Sun, Marilyn Edling of Hewlett-Packard, Dan Warmenhoven of Network Appliance, and Kevin Daly of ATL. Bob Eads of American Protective Services joined Farid as co-moderator: The session went something like this:
FARID: My first question is what are the most significant factors or movements that affected the storage industry in the past year, if you like to comment on their significance and on the way they influence the choices you made for your company. I believe that every one of you want to respond to this. So, Marilyn, if you would start?
MARILYN: The growth in the demand for storage keeps increasing and I think that's a significant change; and then the sophistication of the customers and what you're asking for in terms of a heterogeneous environment; the expectation on the part of the vendors to be able to meet those needs, and also the demands for providing standards choice. I think the biggest change this past year has been the voice of the customer coming through and what the needs are for managing the enterprise storage; all driven out of tremendous increase of information and access to it.
DAVE: I think that Marilyn's comments are all ones that StorageTek would endorse, as well. However, an additional observation would be the number of customers that we have seen over the last two to three quarters that are now opening up their data storage requirement to professional services from us. In the past, we were a vendor that would come in and we would sell our products and then the customer would go ahead and put together the pieces of hardware/software and would put together their own solution into practice. What we have seen as a changing trend is the fact that either the customer doesn't have the time, the resources, and the skills' to be able to do that or they need some assistance. They're looking more for a complete solution and that complete solution could be in the form of doing the job for them, offering them a utility that we talked about yesterday, or perhaps assisting them along the way. So, whereas in the past, we often thought of professional services maybe as something that was only f or the mainframe and for server resources. We've now seen it over the last few quarters change into how do you deploy your storage, how do you deploy your information, how do you get back at your information and use it in the most powerful ways.
MARK: I think for Veritas the biggest surprise was the elusive Y2K problem that seems to have been completely overtaken by the demand of the Internet and the e-commerce market. So we've seen our market's demand very, very strong for storage products of all flavors and all types and that was going through a year that was very, very uncertain, I think, for everybody. The other observation on the last twelve months is that when we last met I think the view of SAN adoption was very, very tentative. I think there's been very good progress made in that area and I think that there's more life in that marketplace and more energy and more momentum in that marketplace than twelve months ago. It bodes very well for the future.
JEFF: So, I think everything we've heard so far, pretty much everybody would agree with up here. I think there's one other thing that hit us the demand for scale. If you look at the Internet today and you just change your perspective, you can't think of thousands and you can't think of millions. You really have to start thinking and approaching billions at the end of the day. And that means that we have to change the way we make decisions. So I think that affects us dramatically and in the decisions that we're making. The architecture that we're building and the deployment of SANs and scaling and growing, I think that's had the most profound effect on everything that we're doing within the organization.
DAN: So, at the risk of being redundant and reiterating everything you just heard, I'd say it's really three factors that I've identified. The first one is the one Mark mentioned which is the explosive growth of online data, whether it be the Internet, e-commerce, or whatever data is becoming more network-centric. The second one is the, I think it follows in the first one, is the separation in most people's minds now of providing a storage solution that's distinct from their computing and applications server solutions. Thinking about that is a separate problem instead of looking for a separate set of solutions. And the third is what the technology has enabled. Thinking back a year ago, I think the most common high-capacity drive was 4GB, and it's now 18, moving very quickly to 36, that's enabled the scale that Jeff mentioned, as well as obviously the Fibre Channel infrastructure, moving away from the limitations of the SCSI connectivity. So, not only do you have bigger drives, you've got a lot more of them an d some very large capacity systems and the combination of those two has enabled people to deploy very, very large systems quickly.
KEVIN: Farid actually mentioned he was going to ask us this question so I dutifully made a list of things which I've been crossing off one by one because someone else has been answering. But I've got one thing left on the list and that's availability. I think the sensitivity to availability in every sense of the term has been enormous this year and is going to change the direction of a lot of the technology and implementations over the next year.
FARID: Thank you. That sets the stage. My next question is in terms of acquisitions, partnerships, and alliances; we have seen an unprecedented wave of acquisitions. What has prompted this? Is this going to continue? What is its significance and is this going to be a differentiator between vendors? I'd like to ask this to maybe at least two people, maybe one who has acquired and one who was acquired. Maybe I'll start with you, Kevin.
KEVIN: Well, I think this maybe parallels a comment that we heard a few minutes ago about following the sun. I think what you see in acquisitions is the tendency to reflect the value proposition in the business. So, as companies begin to evolve their perception of where their value is in the business, their competencies are either going to map well with that or map not so well with that. I think what we'll see is a merging of companies with different kinds of competencies as the business itself changes. So as the value equation starts moving up the systems chain, I think we'll see component companies become more interested in the systems business and systems companies move even further up that chain with higher value of management products.
MARILYN: I would agree one of the things that drives acquisitions is the business strategy. And certainly acquisitions, when done well, can push a company forward. So, I would expect additional acquisitions in this industry. In addition, with new technologies being introduced in the SAN environment, incubator companies become really important to developing new technologies and then, can get absorbed by other companies. So, I do expect to see that continue and then, we have to make sure to do them well.
MARK: I think a couple of things are true. In our case, and I think it kind of echoes the other comments that the mergers that we did were really a function of executing strategy. You build or buy, and the variants you build in are very significant so you end up doing these transactions, which are very risky. So they're difficult to do. But that being said, however, I also think that there's probably now an issue of scale in the industry. And I think that, you know, no announcements of plans on our side but I would say it'd be difficult for a company to establish the kind of scale that needs to be competitive in whatever arena with out doing some consolidation somewhere along the line today. I'd be very surprised if that weren't achieved. So, I think there was an issue of critical mass and scale that's developed in small companies in our rapidly growing marketplace.
FARID: So, what are you saying? If you're below a certain revenue, either you get acquired or you grow fast. Is that? You have only one of two choices, right?
DAVE: Farid, comment if I could. Speaking from a company who in our distant past had done several acquisitions, not very well, these are not necessarily easy things to do. The statistics are 9 out of 10 failed and they failed because the cultures of the companies are different, the synergies cannot be realized, that they always look good on paper, but these are very difficult. Marilyn has some on the way right now, Mark with Seagate, and certainly Sun with Maxstrat and any time you do one of these, they take constant management, so it's not just add one revenue to another revenue and be able to do it. Certainly, another thing to keep into consideration as well, is what it means to the shareholders. When you have a multiple and then you double the finances, you find out that, my God, this is going to be diluted and what not and how am I going to be able to do it so there may be some things that you want to do but when you look at it on paper, the shareholders are going to look at you and say, now you know for this to be created, you want us to wait for three years. What are you talking about here? But that having been said, there are still some alliances and partnerships that you could do without acquisitions. But if you do acquisitions, I think you ought to expect in the future not so much perhaps a software company looking at another software or hardware company looking after a hardware, but truly diversifying their business and reaching out and getting into a totally new business to open up new markets to themselves. And that might be something to watch for.
FARID: Okay. In the same vein, we expected last year to have the networking companies such as Cisco or 3Com to get into the storage business through the storage networking. This hasn't happened. Does this mean anything? Is this important to the industry? Is it good, bad, indifferent ... What are your perceptions of what has happened? Who would like to take this? Jeff?
JEFF: Sure, I'll take the first start for this. I think you're going to see a lot of consolidation over time. I think it's going to take a while. Look at what we're struggling with here up on stage. We're trying to talk about how to do storage area networking, and there's a whole other band of networking infrastructure that's not yet in place and how you connect to the network. But I believe the long run, the big picture, is that everything really will be a utility and that I think the players like Cisco and 3Com are really going to be fundamental to this kind of a technology environment. Then, ultimately you're going to plug in and it's going to work. There's going to have to be ways to interoperate. But I think that we still have to get our own storage area networking strategy underway before we worry about how we plug into a Cisco and make that kind of a switch work `cause the architecture's fundamentally different.
FARID: Anybody has a different opinion?
MARK: My observation is that Cisco, as an example in there, as a proxy for the community, they're engaged in this bandwidth explosion. Their plate is full. The numbers they're talking about, by comparison to storage area networking ...All of storage area networks probably don't add up to 5% of the new bandwidth-based business that these guys are in. I think that they haven't turned their attention to it yet. It's fine for 5 or 10 or 20 companies to go out there and go build the infrastructure and let's just buy`em later, will be my expectation. But I think they're just consumed right now `cause bandwidth is doubling every nine months or something like that and the outlook is to continue to accelerate that.
FARID: Okay. So we should still expect them to come?
MARK: I'm not so sure.
FARID: No? Jeff?
JEFF: I do agree with Mark in the first half. I think they are overly consumed at the moment with the explosion of bandwidth, new technologies like DSL. In Cisco's case, they said publicly they're going to support voice, they're going to compete with Nortel and Lucent and that takes them in the direction away from, if you will, the back end of the computer environment. And I'm not so sure that they'll ever re-enter. But I think the question is time. If they focus on these areas too long, I think they'll find that getting in later will be really hard.
FARID: Okay, that's interesting. Yes, Dave?
DAVE: I think that Mark is onto something here when you look at what are the challenges that 3Com and Cisco has. First of all, 3Com, like all of us, has a core business that they're paying attention to right now and getting that in order for themselves. To move out into the SANs requires some investments, doing some things that may not be a priority for them right now. But I think in the case of Cisco, they're looking at Level 3, they're looking at Qwest, they're looking at Williams, they're looking at IXC and saying there's so much growth out there for our basic products. Why move up the value chain and make that investment at this particular time? They don't need to. They can develop the growth that they have without doing it. You have to understand that even though these are big companies that carry a lot of weight and certainly a lot of influence, getting into this business is not at all that easy. When you think about, well, I have all those hubs out there; I'll just put a few smarts on the hubs; I'll add a cache; I'll begin to manage a storage there and then I'll put a server out there and everything else, you can look up all the gray hair along here, except for Marilyn, of course. That takes a toll. That's a little bit more difficult. It takes investment, it takes a different type of distribution, but I think that we would be naive if we sat up here and we didn't think that they could try if they made the conscious effort to do it.
FARID: Okay. Let's move to another question. Software vs. hardware. We are seeing the software share in the total vendor revenue increase. What is the repercussion in this? Is this important and what is the impact in the channels--resellers and distributors? Yes, Marilyn.
MARILYN: I have a few words to say on that, given our recent acquisition of Transoft. I think that everybody has for information, for access and using it for your businesses and the explosion of different ways of delivering that storage. Unless you have management capabilities for that storage, it's going to become more of a headache than a benefit and software is a great vehicle for driving management. So, what that means is that you're not just going to be adding capacity. It's a much more of a technical solution and as David said, what he's seeing with the professional services, that's a symptom of the customers needing total solutions deployed and that they don't want to just have capacity added. So, software, it's gonna drive a bigger need for technical resellers, systems integrators, etc. So, it would be a change in the channel strategy, as well.
FARID: We certainly heard this total solution thing come over and over again on the sessions here. Does everyone agree with this? Mark?
MARK: I think another important component, and by the nature of the industry, of what software does is that it bridges the heterogeneity gap. Any manufacturer of a system or a device is gonna go and kind of energize his own platform before trying to energize the industry. So, one of the open tasks left for companies like Veritas, a software company, is to help reach that gap in customer where ... I mean, they're dealing with complexity and they want to get it to be simple. They're dealing with unavailability and they're trying to get it to be high availability and they're dealing with it ... all those issues are compounded exponentially by heterogeneity and nobody wants to give up heterogeneity at the customer's side. So I think software is very important in bridging that gap.
FARID: Okay Any problem with the channels? Distribution channels?
MARK: Not for us.
FARID: You've been in the software all along and how about a hardware company? Yes, Jeff?
JEFF: I think the channel is an interesting question because when you ask about software and the channel, you have to get down to, alright, so if you assume software is the value add, let's start making the difference between the backup, remote mirroring, disaster recovery--a lot of the applications that we've talked about--and now you're a channel and you're a reseller. You're going to say so what platform am I going to write that application to. And then once you say what am I going to write that application to, or how am I going to interface it with Oracle, or Veritas, Legato, or ATL? So I think the challenge that software is facing is how do you get to a common interface that you write to? How do you get to the common way to deploy it so that when a reseller uses the software, they don't have to say "well, I'm doing it only on Sun or I'm doing it only on HP?" So, I think it's placing a huge burden on the channel and I think that's what we have to get to and I believe that's what Store X is about. But even outside of Store X, we have to get to a common standard that we write to. Otherwise, the channel will continuously be proprietary in their management and the rest of the solutions.
FARID: Okay Kevin?
KEVIN: I think actually that this is such a phenomenon. In the backup world, that's actually the way it's been, forever. The software applications have pulled the hardware applications into the network so that the relationship between the hardware and the software applications in that particular part of the market has been very software-dominant for a while. I think what's happening now is that it's moving up into a larger part of the storage market and the channels actually have been pretty efficient at dealing with combined hardware/software solutions either by joining the solutions in the channel, or by offering full solutions. I think that if you look at the cost, 75% of the storage cost is people cost.
FARID: Okay Well, we're going to move to another topic, which is the user, and I'm going to have Bob ask those questions.
BOB: As they relate to storage, what are they telling you and more importantly, have you seen any change in the past year or so?
KEVIN: I think that their top priority is survival. There are so many issues we talked about earlier. Tremendous growth in the amount of data they're trying to manage and deploy the movement of the application information from basically a computer foundation into a network foundation. You add on top of that, significant technological change in the area of storage networks and a variety of other things? And why don't we just throw Y2K on top so they have something else to worry about when they go home. And that's a significant set of challenges. You know, I think the emphasis comes back to how do I cope? And Mark said that to a certain degree. You know, there's a number of issues and what they would like to really do is to simplify and in particular to improve their availability, to improve the access to the information to different types of environment and have a solution of skills very easily such that they can handle some of these transitions very, very quickly. So, looking for a survival strategy, I think the traditional approach is of just plugging disks into computers had clearly been shown inadequate to scale to the needs.
MARILYN: We just finished the study in focus groups on different users and one of the things, in addition to the things that were just discussed, is that there seems to be three populations around storage. One is that still sees storage that's traditionally been seen. We call that as sort of the undifferentiated user and then what moves the customer over to more of a differentiated user and probably all of you who are here today in this conference is driven by your application environment. What you need there and that drives the storage technology changes that we've been talking about, the SAN, etc. and the high availability. So, it's from a management perspective is driving the scalability and availability and then there's a third group, which we call the high tech group, which really is adopting the new technology much earlier than the bulk of the rest of the population. So these different groups drive different needs. So, for the middle group, they want simplicity. They want ease of management, the high availability and then the third group is gonna drive the next wave and I bet that's where you'll see the demands for network topology to start changing and a lot of what's discussed earlier today. So, there's different populations and each of you have to understand where you are in each of those groups that drives different requirements.
BOB: And if HP has their way, e-services will also drive up the curve, David?
DAVE: Adding to a couple of things that Marilyn and Mark already said is that the other things that, Bob, people like yourseif tell us, you probably have told us, as well, is that you can't believe how much you're spending on storage. You can't believe how much space it requires and you can't believe every month we're there adding additional to it and so you're coming back and saying, what are some ways that I can save money here? Can I consolidate some of this? Can I use it? Can I share it among my servers? What are you doing for me to keep from having to build another data center? Just to put this all in. So, you're becoming very cognizant of the requirements that storage has. We mentioned yesterday that for many server vendors, it's now over 50% of the invoices they send out are for storage. Everybody is noticing it. So, they're saying, help me become more productive both in the cost of it, as well as how I utilize it. So, these are some of the new things that we're hearing back and so, creative uses of so ftware, creative ways to share data, creative ways to be able to put it on the network, not have to move it to the server resource necessarily all the time. That's what they're pushing for in terms of productivity, as well.
BOB: Right. In fact, I have been talking with some of your folks on that because when the CEO and chairman ask me every month, why am I constantly writing checks for 50 and 60,000 dollars for additional storage, and I try to make the point that you realize that e-mail is doubling every quarter. Our Internet services are doubling every month. Then, they come back and say, well, isn't there an easier way of doing that?
DAVE: And, of course, this year, we're telling you that Y2K, you'd better copy it, you'd better back it up in all of our insidious plan. In fact, people have blamed Y2K is the idea of storage vendors. I don't know if you've realized that yet or not.
BOB: I knew there was a plan under the covers.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Industry Trend or Event|
|Publication:||Computer Technology Review|
|Article Type:||Panel Discussion|
|Date:||Aug 1, 1999|
|Previous Article:||SANs Are Here.|
|Next Article:||THE Automation Challenge.|