Swan Song of a Database Diva

By eweek  |  Posted 2005-10-02

Swan Song of a Database Diva

Habitat for Humanity has a list of skills for potential volunteers to check off: plumbing, electrical wiring, Sheetrocking, roofing, skilled painting, semi-skilled painting, not-so-great painting, etc.

Janet Perna never built a house.

She did build a database empire at one of the worlds largest software companies. Over the past 31 years, shes been one of the movers behind the industry evolution from hierarchical databases to relational databases to a whole industry built around relational database technology.

Shes IBMs general manager of data management solutions for the IBM Software Group, and you could say she was the overseer of the architecture for the DB2 database.

But she cant build a house.

It matters because Pernas retiring, and when youre 57 and youre retiring after a very long and very successful career in technology, you start to think about whats real. And yes, of course databases are real and important and life-changing, and technology is real and important and life-changing.


The best skills she can check off on Habitat for Humanitys list are unskilled painting, and maybe cooking or recruiting. So Pernas going to a Red Cross orientation meeting next week. She wants to go to New Orleans to help out. Shes ready to turn away from autonomic database configuration. Shes ready to turn her attention to walls that are coated with human waste and to lives that are in shards.

Before she goes, Perna recently took time to speak with Ziff Davis Internet News Editor Lisa Vaas to look over her legacy and to say goodbye—and thank you—to the database world she helped to create.

So theres this buzz about you turning up next at Sun Microsystems Inc. or maybe even Oracle Corp. or a startup. Hogwash?

If I wanted to keep working at this pace, Id stay right where I am. I already have the best job there is. But I do want to spend more time in particular with my parents. My dads turning 94. My mom is 87. While theyre still relatively healthy, it made me realize I want to spend more time with them, and they can use some help, so Im going to do that.

Of course. And Ive also heard youre set to do some serious volunteer work.

Ive volunteered with the Red Cross to see if theyll take me to go down and work on the hurricane [cleanup].

Its hard to admit it when youve worked in technology and youve created so much and its sustained you for so long, but I bet that sort of thing does feel more real at this stage of your life, doesnt it?

Its the return part of life. The first part of your life you learn, then you earn, then you return.

Lets talk about your legacy. And lets talk about where you think the future of database technology is heading: where it will be in the coming years, after youre gone.

Ive been going through my files, so I had my own journey through memory lane.

IBM has been in the database business for close to 40 years. Ive been at IBM for 31. Weve moved from hierarchical databases to relational databases to a whole industry built around relational database technology.

When you think of all the companies that were there, in the days of the big database wars, with Informix and Sybase and [when] even Borland had a relational database, before Microsoft... We had Ingres. … And now whats happened over the last five to six years is a charge IBMs been leading around an information infrastructure. With a set of data services and content management services and information integration services, and with IBM having made a number of acquisitions, Ascential being the largest.

Click here to read about IBMs expanding SOA management practice.

And then thereve been things like data models for banking, insurance, retail, and the whole area of master data management and acquisitions [such as] DWL.

So when we look at information in the future, information will be presented as services, as part of SOAs [service-oriented architectures].

Were certainly hearing a lot about SOAs and middleware, between the three big vendors in that area: IBM, Oracle and SAP. But thats not near future; this stuff will take a very long time to construct.

I think it will be evolutionary, I agree with that. It will happen bit by bit over a number of years. But will it be any more painful than anything else weve been through in the last 10 years? I dont think so.

But implementing large monolithic applications, some could say it was pretty painful, but in reality they were an improvement over a lot of homegrown work that was there.

So would you say were at the verge of another evolutionary leap?

There are different points in time to view modernization projects. In 2000, remediation drove modernization of systems and installation of packaged applications like ERP and CRM. Companies said, Ill replace what I built 20 years ago with these packages and standardize on them.

Was that painful? Yes, but they had to do year 2000 remediation anyway, and they modernized along the way.

Well transition again to get to SOAs, where well make revisions to business processes and well build these components on a new base in a new modular approach to doing that as we modernize these systems.

It wont be an overnight, everything-is-brand-new scenario.

Looking back, how much of a difference did IBMs embrace of Linux make to IBMs business? Looking forward, where do you see IBM and Linux going in the days ahead?

Where IBMs embrace of Linux made a tremendous impact was on the acceptance of Linux in the industry as a server platform. Linux clearly began to be recognized as a mainstream, commercial platform once IBM embraced it.

IBM also has over 1,000 people working in the open-source community around Linux and making Linux more scalable, more industrial-strength.

Next Page: Oracles Linux lead and Project Fusion.

Oracles Linux Lead and

Project Fusion">

But at this point, Oracle has a growing lead over IBM in this subsection of the market, with growth of 155 percent, according to Gartner. The Stinger database was optimized for the Linux 2.6 kernel. Whats the uptake been in the Linux market, and will Stinger help IBM in this market section?

When you look at the whole size of the database market, what was published was that Oracle had about $600 million in [Linux] revenue. Look at what Oracles been running, and youll see they were pushing their Unix customers to Linux.

These were not new customers; they were Oracle Unix customers. Oracle was moving them to Linux, and selling RAC [Real Application Clusters] on top of the Oracle database. Thats where the revenue was coming from, and thats where the tremendous growth came from.

At the same time, Oracle had a decline in Unix [revenues] and growth in Linux. So what they did was displace some Unix with Linux, and sold RAC along the way.

Read details here about how integration between Oracle and IBM products may shake up the applications market.

And they ran that play very successfully. What you saw for a year was double-digit database growth. It just wrapped around, and they just announced earnings a few weeks ago for the quarter, and database business slowed. You say, What happened there? Id say their wraparound play has wrapped around.

They sold their RAC, they moved their base, and now they have to grow the old-fashioned way and get new customers, instead of leveraging their incumbent Unix base to grow on Linux.

For IBM, we havent had the enormous Unix base Oracle has had. Any new sale on Linux is a new customer for us.

We grew at double-digit growth last year. On Linux. For us, any customer is a new customer.

Let me ask you about the apparent Oracle-IBM rapprochement, with Oracle shifting to this new stance of supporting WebSphere in Project Fusion as they tie together all the technologies theyve acquired this year. Is that significant, or is it marketing hyperbole?

Whats happened here is Oracles customers, their application customers, are saying Were committed to WebSphere. Were committed to WebSphere as our middleware platform. And Oracle has come to the realization that an application strategy that doesnt have strong support for WebSphere isnt going to work. Its a loser. Customers have demanded that. Customers have voted.

Now Oracle, if theyre serious about the applications business, which they must be seeing as how they spent billions on it, to meet customer requirements, are going to have to do a very big job of supporting WebSphere, because that is the infrastructure.

Theyre hearing also on the database side that DB2 is a standardized platform for them, a standard part of the infrastructure, and youre starting to hear Oracle saying, We will continue to support DB2 with their Siebel and PeopleSoft and JD Edwards acquisitions, and we may support DB2 with Fusion as well.

As the market votes, as the customer votes, Oracle is noting more and more how prevalent WebSphere is.

Looking back, what are you really proud of? The Informix acquisition? The self-managing database?

Thats a tough question over a 30-year period. The thing, the most sustaining achievement, is the team you leave behind. Its the people. The people really make this business what it is. When I look at it I look at the technical team, the sales team, the executive team that Im leaving behind that will carry this on, not only on behalf of IBM but on behalf of customers and partners. Thats truly the legacy: Can what youve built sustain itself?

Im very proud of the team, the organization, Im proud of the business weve built together. Weve taken it from a business of less than $1 billion to a multibillion-dollar business. Weve significantly increased the size of the team. And moving from a database technology to where we are today with information integration, content management, and master data management, to all the acquisitions weve done, and our ability to integrate all the companies weve acquired through the years.

Thats tough from many perspectives: from people coming in, to people who were here, who continue to embrace new teammates, new colleagues. For me its about the people who are here and their support and the culture of dedication.

That Im the most proud of. And that does result in good business.

At the end of the day, thats what its about. Thats what it is. And the relationships Ive been able to build with customers and with partners.

At the end of the day people buy and do business with people. Transactions are done between people. You want —particularly in an area like database technology, these are bet-your-business decisions in a lot of cases. And you want to do business with people you can count on, and whom you can trust and who will be there and who are dedicated to your success.

Everybody can say that, but not everybody can deliver.

Check out eWEEK.coms for the latest database news, reviews and analysis.

Rocket Fuel