Midmarket Momentum In Storage

VB: The word on the street is that the storage midmarket holds the most opportunity right now. What are some of the functions that have migrated from the high end to the midrange?
Canepa: First of all, it's the functionality of the system itself. Everything is redundant. Everything is hot-swappable. In the past, you saw these capabilities only in the expensive, bigger systems. The fact that you can repair the product without having to take down the application%85that kind of capability is finding its way into the midtier systems. In the past, features like Snapshot were found in big, expensive solutions. Other examples include phone-home capability and network-system management, as well as home monitoring. These were once the province of big, expensive, million-dollar machines. Now midtier machines have tools that allow you to proactively discover a disk drive that may be nearing the end of its life.

VB: What are customers asking for? Do you see yourself guiding them to the midtier?
Canepa: Increasingly, customers are speaking the business language. They talk about the total cost of ownership of their storage environments. For every dollar of storage drives they buy, they are still spending $8 or $9 to just manage it. What they are very much interested in is reducing those numbers. Second, they talk about the overall total cost of recovery and total cost of disaster recovery. With everything that happened in the past two years, it is becoming a bigger thing in people's minds. E-mail, for instance, is becoming a regular issue. You need to keep e-mail for long periods of time, and so companies are under the burden of making a lot of these things happen. They want to be able to comply at the best possible price. Our response is to just continue to drive this robust data-center functionality further down the product line.

VB: So what does this ultimately mean for the high-end market?
Canepa: We think there is always going to be a high-end market. If you look at the server equivalent, there still is a mainframe market,even though the vast majority of computation is done with big servers at a much lower total cost. If you look at the whole Unix industry, the same thing happened. We started to build midtier boxes. Fifteen years ago, people thought that a Unix machine would not be able to compete with the mainframe. There was a big gap. But over a few years, we started to build better Unix servers, and we drove data-center functionality into the Unix servers.

Pretty rapidly we could do the vast majority of data-center server applications, at least as well or better than mainframe.

id
unit-1659132512259
type
Sponsored post

I think the same thing is happening in storage. The traditional, monolithic storage box is going to get con-fined to a particular set of applications where customers want absolute big, monolithic design. But for more and more applications, these midtier systems are going to be far better solutions.

VB: When you go into a high-end competitive bid, who do you come up against the most?
Canepa: It depends. If the customer is simply looking at storage, then EMC tends to be the traditional competitor there. If it's an end-to-end solution, where the customer is looking at servers and storage or buying a complete infrastructure environment, then, by far, IBM is the typical competitor.

VB: Let's go back to the midtier. Tell me where Sun is investing in the midtier. Talk about your products and road map.
Canepa: We have been investing in the midtier for about three or four years. We have a midtier line of products. We have the 3910, 3960, 6910 and 6960. They are basically able to scale to about 20 TB, which is a pretty good size for a midtier system. We are able to offer significant price savings on a per-megabyte system compared to other systems. We are putting the final touches on a total refresh in the midtier product line. So you need to watch this space because we are not that far away from showing new systems and technologies.

VB: On the midtier side, which vendor do you often come up against?
Canepa: In the midtier, there is a bit more of variety out there. Certainly, we see the [EMC] Clariion quite a bit. We see [Hewlett-Packard's] EVA every once in a while. I think HP in that middle-tier system tends to stay close to its installed base. So we will see HP in an account that tends to have a lot of HP servers, or we are sharing a SAN and both of our servers are on it. As you go further down the line, on the lower end of the midrange, we tend to see more of the Compaq EMA line. We tend to see them in the channel.

VB: Has EMC become a more significant force with Clariion considering its partnership with Dell?
Canepa: We think the partnership with Dell has allowed it to go into the channel primarily to attach storage to Win2K and T-base systems. Right now, by and large, we see that Wintel SANs tend to be separate from enterprise SANs. We don't see Dell having the capability, the professional services or the ability to write support contracts to really come in [and put a mark in the enterprise space]. It tends to focus on the NT space.

Customers in the enterprise storage environment are a pretty conservative set of folks. So selling storage is not about selling disk drives. Selling storage is about selling end-to-end capabilities. It's about selling support, guaranteeing that the data won't go away, helping them assess their storage utilization, and giving them software tools or professional services to help them architect, implement and manage their systems. We do millions of dollars a year as a company in managed services for clients. A lot of those are storage-managed services, where we do something as simple as monitor their systems, or more sophisticated managed services. Those are things that companies like Sun can do. Dell is a long way away from that.

VB: There has been a lot of talk about storage-management software. Is it still considered the market that everyone is anticipating it to be?
Canepa: I think a lot of differentiation is going to be rooted in software. But, we think, over time customers are not going to go out and buy software in the traditional sense. We think, increasingly, software is going to be an inherent set of functionality that comes with the storage system. We believe very much that customers want to buy systems, and systems are going to have functionality.

That functionality is the result of a bunch of engineers writing a bunch of code. But it won't get marketed as software, especially from systems makers. It will get marketed as features of the storage environment. So it may look like hardware, even though 70 percent of the engineers I have are software engineers. Most of the functionality of the storage system manifests itself through software engineering.