Backup for the Community,
Present and Future

Paul Meszaros leads a company that backs up a 3000 customer’s past, while making connections with the future. It’s a familiar position for a long-time MPE vendor like ORBiT Software, with thousands of customers around the world. The majority of the company’s business remains in the HP 3000 marketplace, but about half of those customers are trying to decide what to do about their platform’s future. ORBiT and Meszaros, always a good touchstone about the community’s intentions, want to provide choices while they protect what’s working.

Meszaros came to ORBiT in 1993, after more than a decade of work in the healthcare and computer industry. He had brought a biology and chemistry degree to his drive to become a doctor, but landed in healthcare’s computer commerce instead. He met his first HP 3000 in 1981 at an aviation company where he was selling a turnkey system. “It was as big as a refrigerator,” Meszaros remembered, and the 3000 wasn’t displaced by the Data General system he was selling that day.

After selling managed care systems through Computer Sciences Corp. in the early 1990s, Meszaros joined ORBiT in 1993 as general manager of a company with a worldwide installed base of customers using HP 3000s. ORBiT made him president eight years ago, and he assumed responsibility for the company’s US, French, German, and UK operations. Under Meszaros’ guidance, ORBiT has become a place where third-party packages come to be sold, as well as a development house that breaks ground connecting MPE customers to the latest backup technology.

While ORBiT and backup are a well-known match, the firm has spread out into a broader datacenter solutions company, too. We wanted to ask Meszaros about what his customer base is thinking about its transition, and how the latest backup technology is being adopted by the 3000 customers. We spoke in the week before the West Coast Solutions Symposium in late March, talking about future destinations for a community still looking for ways to get there.

Linear Tape Open-2 backup devices are out there, and prices are falling. How can customers use these with HP 3000s?

There are a number of powerful tape devices on the marketplace which are not supported on the HP 3000. It’s becoming increasingly so as manufacturers back away from offering High Voltage Differential devices, which are required if you’re going to use them on any 9xx Series HP 3000.

Our solution to this is a new product, which we’re going to roll out very shortly. It enables an HP 3000 user to use any tape device in the world, made now or coming out. It’s a software solution with a hardware component in it, allowing an HP 3000 user to use LTO and SuperDLT libraries, as well as all other devices on a network. We call it the Virtual Storage Engine.

This allows HP 3000 administrators to take advantage of resources they might not otherwise be able to use. We like that because it extends the usefulness of the 3000, and removes barriers that might be in the way of fully utilizing the HP 3000 in a distributed environment.

Why does LTO offer to the 3000 user?

There are some drawbacks to DLT architecture, like the way the tape meets the tape head. This newer technology eliminates this complication, reducing a potential failure. LTO also offers greater capacity than DLT.

Is HP still responding to service requests and enhancement requests that relate to backup devices?

We’re on the outside looking in at that process. I will say that when HP does these patches, sometimes they are pull releases, not released to the general world. The biggest problem we’ve seen in the past year is that when new patches are released, customers aren’t participating in beta testing.

What advice do you have for customers about their 3000 backups now, in an era when lots of companies are making plans to leave the platform? Should they be backing up to non-MPE systems?

People are going to have face up to the fact that at some point they’re going to have to commit to homestead on the HP 3000, or go to another platform. Many people who are on the 3000 who haven’t committed to homesteading are still uncertain as to what platform they’re eventually going to land on. Being able to back up their HP 3000, knowing that whenever and whatever they go that they’ll have access to their historical data, would be a reassuring thing. We offer our Rosetta Store solution; it allows that data to eventually be restored to non-MPE systems. In the medical field, new HIPAA regulations say a patient has the right to know about the disposition of their confidential information, and they have the right to ask six years into the past. If you’re a big HMO and you’ve got 500,000 members, you may not want to keep six years of data online. You either have to keep an HP 3000 cooking in the back corner or you’ve got to make arrangements to read that legacy data on your non-MPE machine.

What do your support renewals tell you about the stability of the 3000 customer base, in general?

Over the past two years we have been tracking this very closely. Early on, we estimated that the installed base would diminish by 20 percent a year. That hasn’t happened. The first two years it was closer to 11 percent a year. This year we expect that to accelerate a bit, and we’re trending at about 15 percent shrinkage. However, we are still adding new customers, so the net decline is somewhat less.

I don’t know whether the fact that October, 2003 has come and gone has accelerated anything. A survey that we did recently showed that 22 percent of our customers indicated they were going to homestead.

How are the ORBiT migration tools being accepted so far? Have Rosetta Store or Cobol2C helped out at some customer sites?

Frankly, the bad news is that the migration wave hasn’t really started yet. Of course, that’s also good news for us. We’ve seen interest in both Cobol2C and Rosetta, but we haven’t reached that point where the migration plans for a majority of companies are underway, and they see the immediate need for those products. I think we’re still a little ahead of the curve.

Are a serious share of people still deciding what to do about their Transition?

Exactly so. When we did our survey, about an equal number of people as the homesteaders, 25 percent, had made or were executing their migration plans. The remainder were undecided.

Do you think the customers’ indecision reflects how difficult is has been for everyone to plan?

Yes. There’s also a certain amount of animosity we see; people are angry at HP for letting them hang out in the lurch.

Some MPE companies are really starting to shift their focus to software they offer on alternative platforms. How much of Orbit business still lies on HP 3000s?

Currently, our MPE business is still our bread-and-butter business, the majority of our business activity. There’s no getting away from the fact that the marketplace is going to diminish. Years ago we faced up to that fact, even before HP made its 3000 announcement. We weren’t necessarily hedging our bets about the future of MPE, just trying to better address the needs of our customers, who over the years have started to deploy more open systems in their environment. We’re no longer just an MPE shop. We look at ourselves as being in the business of ensuring data protection, data storage, data management. That’s why we got into things like job scheduling, output management and virtual storage.

In one way, we’re obviously committed to our HP 3000 customers for as long as they remain on the 3000. As our customers broaden their horizons and look for other non-MPE solutions, we have solutions and services for them in those areas, as well.

Do you have faith that the homesteading marketplace will provide revenues in the years to come?

Our indications are that it will. Obviously not to the extent that it did in 1998, but if we believe what we hear, there is a significant number of companies homesteading.

And there are some surprises in that group. A very large company has gone through their due diligence, and examined how much it would cost them to move, specifically to HP-UX. The cost they estimated was $50 million and wouldn’t even get them anything new: no more functionality or power. When you add $50 million onto your cost, in the industry this particular company operates in — a market with very small margins — it’s a no-brainer. They just can’t spend $50 million to be in five years exactly where they are right now with HP 3000s. They just closed up the issue and maximized their position on the 3000, to make sure they can operate indefinitely on what they’ve got.

You’ve got cross-platform offerings like Spinner and Dollar Universe. How do you qualify these kinds of packages for your product lineup?

As we consider other products to offer, we’re trying to find those that are relatively seamless as far as their applicability on MPE boxes and open systems boxes. We can offer solutions to a company that is saying today, “We’re not homesteading, but I don’t know where we’re going, and I don’t know when.” We’re able to say, “When you get to where you’re going, you’ll already have software that is transportable to your other platform.”

Is the 3000 community using networked storage to a significant extent?

I don’t think so. Networked backup is relatively cumbersome in terms of performance. That was one of the areas we attempted to address with the Virtual Storage Engine. The 3000 uses it to see virtual tape drives, as many as you want. You do a store to all of them simultaneously, and it takes the bottleneck out of that procedure.

DDS-5: Necessary evil, or improvement that makes DAT useful at last? Support it with Backup?

It’s a continuing evolution of DAT, and any time you’re getting more performance it’s a good thing. But in my opinion there’s much more powerful technology available for tape backup, viable options in all but the smallest environments.

DAT is still the common denominator of tape technologies. If you wanted to be safe, it’s almost like using Microsoft Word to exchange documents. If HP supports DDS-5, then we will, too.

What’s your view about HP electing to support 6.5 for an extra two years?

That’s a good idea. Frankly, I think anything that HP does to assist the 3000 community to be able to use the equipment for as long as possible is a good thing. It’s no burden at all for us to continue 6.5 support in our products.

Is an MPE source code agreement between OpenMPE and HP essential to the 3000 market’s survival?

For the MPE market to survive, it’s really got to grow. It can’t stagnate. And if it’s not growing, it’s probably shrinking. In order to be a growing, viable market, there has to be an ability to create new licenses and sales of MPE. From what we’ve been able to see in the last two years, I don’t see progress, concrete steps being taken. I don’t see HP and OpenMPE coming to agreement on difficult issues.