News Stay informed about the latest enterprise technology news and product updates.

Executive vision: Compaq chief digs in his heels

In the first in a series of interviews with the CEOs and top executives of major storage companies attending the Merrill Lynch Storage Technology Conference in Santa Barbara this week, searchStorage talks to Mark Lewis, the newly-named general manager of Compaq's Enterprise Storage Group. Compaq Computer Corp.'s enterprise storage initiatives have been hindered by the popular perception that a company with a PC desktop heritage can't possibly be serious about corporate data centers. But Compaq has come a long way. Recent research has pegged Compaq as the fastest growing storage vendor of the last two years. Gartner Group recently awarded Compaq the coveted upper-right quadrant in its ratings of storage-area network vendors. And Compaq is now rolling out an ambitious strategy to raise a single software and management umbrella over customers' entire storage holdings, a strategy it calls virtualization.

Would you say your name recognition is where it needs to be?
It's on the right track. I think in most circles we are considered the storage alternative. We have a phenomenal success rate when we engage [the competition]. Contrast your strategy to EMC's.
We contrast ourselves by proving to customers what we can do with more modern, modular architectures and the raw price/performance we can provide. We tell them they don't have to give up any features. We just save them a lot of money. And we're not ashamed to say that. We have a very strong vision around virtualization. We believe we can show customers how to buy a SAN today and protect their investment. How much of this business is outside of the traditional Compaq/DEC/Tandem base?
The whole storage networking thing has kind of muddled that. We have a good amount of non-Compaq equipment in the SANs we're building. I would say over half of our customers have some non-Compaq servers and storage in the mix of their SAN implementations. You're having to make a transition from being a hardware provider to a software provider, which Compaq hasn't done in the past. How are you going to get there?
We are the fastest growing major vendor in the storage arena today and about the eighth largest in total volume for 1999. Our growth in 2000 was about 200%. We are more than pleased with our progress. But culture and mindset are the slowest to change. I can change our product set a lot quicker than I can get the guy on CNBC to stop calling us a box company. How do you plan to build out this virtualization message so your customers understand it?
Virtualization is not due out till the end of this year so we have to build understanding of the value. Now that we've announced the VersaStor technology, it's a little easier. We're doing the product with a lot of partners, including IBM and about nine different partners who are going to develop the virtualization products with us. We intend to do a lot of shows and events and spend more time talking about it. And we will build value models to show customers exactly what can happen when they buy the product. It is all about return on investment, not gigabytes and MIPS. It's all about space efficiency, downtime costs and flexibility. The virtualization market will have to go through its shakeout period. It's very hard to tell who's going to be well aligned with the virtualization trend overall. We try to partner very well, even with companies, like Veritas, where we have some overlap. You mentioned that you're going to overhaul the whole product line this year. What form will that take?
There'll be a new generation of controllers that will work with the existing products but will be faster, better cheaper. We're putting in server-less backup agent technology in the SAN. We'll roll out virtualization and we'll continue to refine the management tools for storage utilities. Our goal is to not run storage utilities but to let our partners do that with all the software and hardware tools they need. So you wouldn't rule out outsourcing your entire array business?
That's many years away, but we wouldn't rule it out. If the market became truly a commodity there, it's a possibility in the very long term. But we think that market is good for some time to come. But we have data routers, switches and host adapters, a lot of things that are closer to [becoming commodities] than the RAID arrays. Last year the market was talking all about NAS, this year it's all about SAN and now you've got virtualization on the horizon along with a host of new connection options. Is there a risk of customer whiplash?
Absolutely. We are scared about customers getting fearful of implementing a network storage strategy because of all the buzzwords and the change. However, we don't think it's changing that rapidly. Customers look for a protected investment from us, where technology changes like Infiniband or iSCSI become very transparent to them [if they become important]. Having said that, I think virtualization is an exception that has very direct customer impact. That's because it becomes another value layer on the SAN. And the pure cost savings it can provide make it more of a business proposition. Many of the other technologies are just about bits and wires and we try to shield customers from that. If virtualization is successful, doesn't that render the devices commodities?
Everything will trend that way and you have to continuously innovate to maintain your growth and margin. We do not intend to ride the plane into the ground. Almost all of our new growth in storage will be in software and services. The same applies to our investment fund. But things that are the lowest value to customers we may move from an internal to an external source. What is the status of the PCI/X technology that's being developed by Compaq?
I can't tell you much about the specifics, but we follow the lead of our server divisions and make sure we have the right storage attachments. PCI/X and Infiniband are all going to be okay. The network attached strategy for Compaq is branded today as the TaskSmart server that's being jointly developed by the server and storage groups. Our approach on NAS is that it's complementary technology that should be implemented as part of an appliance strategy. That means you can put storage directly on the NAS appliance, but as your storage requirements grow, it will exist as a front-end appliance to the SAN. This is another differentiator from the major competitors. Some folks are bringing out NAS product lines but not tying them into the SAN. Other people are saying NAS is the way to go, but they don't have any SAN at all. We believe those strategies are shortsighted. What do you make of the Fibre Channel vs. IP debate?
We believe Fibre Channel is here and deployable today so we see no reasons for customers to shy away from Fibre Channel SANs. As iSCSI develops, it may be a complementary or a replacement technology. We will make sure customers' investments are covered. Where are you going with your investments in other storage companies?
We've made major investments in three companies: StorageNetworks, HighGround Systems and Precept Software. We have about $100 million we'd like to invest this year and we're going to focus on two areas: storage software and service providers. We want to stay on the value side of the business. We are staying in the high end of the value chain. We aren't out to buy out a bunch of other products and resell them. We want to innovate in storage--and less so in the hardware area [than in software]. You say you're the alternative to EMC. Can you foresee a realistic scenario where they could be the alternative to you?
For many customers today, that might be the case right now. We are the leading provider of storage in the whole Windows NT market. And, we shipped last year between 70 to 80 Petabytes in total storage, depending on which research you look at. The most recent IDC research says that's more storage than EMC and Sun put together. So if you look at real capacity shipped today, it might be a closer race than one would think. Of course, our competition sells at the very high end for a lot more money but not necessarily for a lot more capacity. We're selling the same functionality but our list price is a lot less. It's generally accepted that the cost of management is five to seven times the cost of hardware. How can that be improved in a well-managed SAN?
I don't know that SANs can improve that ratio substantially becomes there are too many other variables in that equation. We think virtualization can double the amount of storage that can be managed by a single administrator. It can provide a 50% improvement in utilization. In other words, if you have 10 Terabytes of distributed storage deployed, you're probably using only half that now. By putting it in a SAN, you can probably make good use of 7.5 Terabytes of that. Through virtualization, we think we can take that to 9 Terabytes.

Dig Deeper on SAN technology and arrays

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.