Hewlett-Packard Co. (HP) is "at a loss" about Aperi, IBM's open source storage management group. That's how it was phrased by Frank Harbist, vice president and general manager of information and lifecycle management and storage software in the StorageWorks division at HP. Meanwhile, Ash Ashutosh, erstwhile CEO of the newly acquired AppIQ Inc. and current chief technology officer of the storage management software group at HP, had plenty...
to get off his chest concerning what he called "misconceptions" about the SMI-S standard.
This talkative twosome also found time to fill us in on their mostly completed merger, as well as what other technologies HP might look to add next.
What is the difference between IBM's storage management initiative, Aperi, and AppIQ's storage management initiative, OpenIQ?
Frank Harbist: We don't really know enough about the details. But from what we do know I think they have gone too quickly directly to an implementation than focusing through the standards committee to finalize and ratify the specification. What we've learned is that the information that has been shared has moved to an implementation and business model before closure on spec, and it's HP's view that that's not the right order. We need to work on the spec and then sort out the business model and implementation.
So what's the status of the OpenIQ program? How many members does it have? What is it specifically working on?
Ashutosh: This is the biggest confusion. OpenIQ is not a program signing up members. We get asked, 'Well, how many members do you have? Aperi has this many.' But for us it's a call to action to get the open standards body to build the next generation of SMI-S.
What do you believe is the ultimate future of the SMI-S standard?
Ashutosh: There's a roadmap. We're already working with Version 1.1, which includes performance enhancements. Virtualization is probably the next step. There's also going to be a move, along with the rest of the industry, to provide higher level management of storage utilities. We're trying to make the end user's management problem easier one problem at time, starting at the bottom of pyramid. There's a hierarchy of needs.
It's been said that the industry actually fears true heterogeneous storage management, since it would blur differentiators between products. How do you reconcile selling a heterogeneous management tool with trying to sell your hardware?
Ashutosh: The most tangible example is what AppIQ has already done with one of our OEMs, Engenio [Information Technologies], creating a very specific performance analyzer module. There's a general performance analyzer module we sell for all storage devices, but for this particular OEM, who has marketed different high-performance computing products, we've been able to go back and dig deeper into performance capabilities. Another one of those myths about heterogeneous environments and standards-based tools is that everyone goes to the lowest common denominator. For most operations, you do want to provide a basic set of operations on a day-to-day basis, across platforms. But SMI-S does allow us to expand into those places where a product may be different from rest of the market.
It's true that with home computing, you don't think about buying Linksys router anymore -- you just pick it up at Best Buy. I think enterprise is headed there eventually, but we're a long way from there still.
Harbist: It's an absolute reality for us that companies that don't compete within heterogeneous storage management will not succeed. There are still ways to set yourself apart whether it's customer relationships, or a total solution orientation, etc. But we see the storage market becoming a more integral part of the general IT market and we feel that the IT players who take steps to bridge across domains will be the winners.
What was the vision behind the acquisition of AppIQ in terms of the benefit to storage users?
Harbist: About a year and a half ago we created a strategy to try to bring together server and storage management, and to simplify the IT environment. As we embarked on that, we found a very good match with where Ash and the AppIQ team were going. There's a focus for them, too, on rich functionality, ease of use and an orientation toward standards connecting into broader management paradigms. We partnered with them and immediately began receiving great feedback from customers, and felt like we were on to something. From there it was a no-brainer to us to formalize it and to take advantage of the intellectual property.
What steps have been completed in the integration process already?
Harbist: From a product perspective, we've released System Insight Manager (SIM) 5.0 and Storage Essentials (SE) 5.0, which we see as the bookends, if you will, of the eventual product suite. The SIM platform combines server and storage management; Storage Essentials adds modular management tools. We've also integrated our field organizations; we've created in each of our regions responsibility teams driving business for HP. We've firmed up OEM business models and dedicated teams to support Sun Microsystems, Silicon Graphics Inc., Engenio and other OEM partners. Finally, AppIQ is physically here in Marlboro [Mass.] in an HP facility -- that happened just a week after the acquisition was closed. From here it's the minutiae of acquisition -- getting everyone HP business cards, things like that.
How do you plan to gain revenue from AppIQ?
Harbist: The revenue model is essentially the same revenue model as we have today. Customers get the unified information management framework via a basic configuration of SIM for free -- you can download it off the Web and it's on every HP storage device they buy. For a lot small and medium size customers, that's more than sufficient. But for large enterprises, it's not -- that's where we can offer them Storage Essentials, which is basically a set of plugins. The Essentials family is modular and not for free. Customers can buy what they need for one particular point in time and add as they grow. There's significant value there just because of the sheer number of HP servers going out with Insight Manager already installed as the framework for going to Storage Essentials -- HP produces a ProLiant server every 26 seconds. So there are lots of systems out there with SIM that will allow us to connect sever management and storage management capabilities in the future.
What convinces you enterprises will want to go above what's already included?
Harbist: It's been our experience in terms of how they're manage current environments. For example, the basic SIM offers some discovery and capacity monitoring. Enterprise customers want to do an awful lot more than that -- provision -- so the Provision Manager Essential fits into that, and there are other management modules for Oracle, SQL, Exchange, etc. They're going to need those.
What is on your roadmap for future development?
Ashutosh: Something that combines more components of storage, including offline storage, NAS, virtualization, grid computing … there's always new technologies, and we want to make sure all components can be supported in our environment. There's always more integration to do. It falls along three dimensions: supported devices, proper functions in storage utilities and tighter integration across domains and platforms, between server utilities and storage.
HP dropped its own proprietary SAN management software, OpenView Storage Area Manager (OVSAM) in February, in favor of reselling AppIQ's SMI-S-based Storage Authority product line. You said you were migrating some 4,000 OVSAM customers to AppIQ's software, over time, based on user preference. How many of those users have migrated to date?
Harbist: We have not obsoleted the OVSAM product and don't see it happening for a while yet. We continue to get orders and enhancement requests from customers. There's an ongoing revenue stream there. But we think the acquisition of AppIQ will accelerate the transition because customers will have more confidence in HP backing the technology itself. Once we end of life a product, we continue to support it for five years, but we haven't gotten to that yet. We could probably expect end of life within the next year so then there's a six year clock of the support afterwards.
So out of those 4,000, how many are AppIQ users today? A majority?
Harbist: I would say no. Backup environments are hard to change. That plus our transition plan, which didn't force or pressure customers into making the switch. So some customers are piloting the new products, some are partially integrating. We're letting them work on their own terms.
So you expect the rest of the customer base to migrate within a year?
Ashutosh: Remember, it's been less than a month since the acquisition was complete and already migration has been going well. It takes a little while to ramp up.
Harbist: Having gone through the Compaq merger, we're experienced at this. Our absolute priority has been that whatever we do, we do it on the customer's timetable and work with them so they understand the future direction while we support what they have in place today, as long as practically possible.
In terms of future acquisitions, what other areas of the storage sector look interesting that HP does not play in today?
Harbist: I do see a new paradigm in data protection, which will combine replication, traditional backup and archival processes. Another is moving from storage management to information management -- content containers. It's not about blocks of data; it's about the information and how everything relates to each other.