Chapter 6
The Integrated Database

Businesses today are under intense competitive pressures. All of us are being asked to do more with less and to do it faster than ever before. In our fast-paced world, standing still is a sure way to fall behind. As Yogi Berra, that great American philosopher, supposedly once said, “It gets late early out there.”

In this competitive environment, a great untapped resource for many businesses is their operational data, which they use to run the business. In fact, many businesses now realize that their operational data may be their most valuable asset. This data is typically retrieved, updated, and stored as part of an application program, such as an accounting application or an order-entry application. We usually call this type of application online transaction processing (OLTP). With OLTP, the application often is interactive, whereby the data is updated through online transactions. Businesses typically store operational data in a relational database for easy, fast access.

The digital revolution has made it easy for businesses to capture and inexpensively store lots of operational data. In addition to storing operational data because it’s easy to do, many businesses store their historical operational data because they believe valuable treasures are hidden inside. But then these businesses frequently ask, “We know that hidden inside our data is information about critical markets, our products, and our customers, but now how do we get at the information to make intelligent business decisions?” Traditionally, data analysis to find the hidden treasures was strictly a manual process. But in the past couple of years, businesses have begun to recognize that automating data analysis has become increasingly cost effective. We often call this data-analysis software a decision-support system, and we talk about transforming our data first into information and then into knowledge.

Data warehousing and data mining, words unheard of just a few years ago, have become part of our everyday business vocabulary. Stories in the press about companies that have successfully employed these technologies to gain a competitive edge in their businesses have motivated other businesses to start down this same path. Implementing a data warehouse, where the information gleaned from the operational data is stored, may be the fastest growing application in business information technology. Some consultants have predicted that fully one-third of all information-technology spending in the next couple of years will be related to data warehouses.

What does all this have to do with the AS/400? More data warehouses are likely to be implemented with AS/400s in the next couple of years than with any other system. The reason for this is simple: The AS/400 is where the business data is located.

DB2/400, the AS/400 database, is the most widely used multiuser database in the entire world. DB2/400 databases contain more operational business data than does any other kind of database. This information usually comes as a surprise to many people, including many AS/400 customers. IBM doesn’t sell DB2/400 as a separate product, so most people in our industry are not aware of its position in the market. They often guess the most widely used multiuser database comes from Oracle or Sybase, or from Microsoft. Very few of them know about DB2/400.

With so much business data in DB2/400 databases, it stands to reason that many AS/400 data warehouses will develop. We at IBM estimate that more than 60 percent of all AS/400 customers will implement some form of decision-support system in the next couple of years. In the following sections, we look at this increasingly important role of databases in business and at how the AS/400 is positioned to be a premiere player. We then look at where the DB2/400 database came from and how it is integrated into the AS/400. But first, what is DB2/400?

The Database with No Name

A long time ago, we decided to integrate a relational database into every System/38. That idea was carried over to the AS/400. We reasoned that a high-function relational database capable of efficiently and reliably handling large volumes of data was the foundation for all good business applications. Rather than making the database a separate product, we just built it in. We also didn’t bother to give it a name. Under the covers, the anonymous database quietly did its job for many years without much fanfare. As a result, many of our own customers have not known they have had such a powerful database working for them.

Some surveys taken a few years ago showed that as many as 40 percent of our customers were unaware that they had a database and that they were using it in their business every day. This is both good news and bad news. The good news is such customers don’t do anything special to manage their database. They just use it. They certainly don’t have anyone identified in their business who has the role of a database administrator, unlike some other databases that require people to manage the database.

The bad news is that many of our customers thought someday they would have to add a database to their AS/400. All modern computers, even PCs, have relational databases, which the user purchases separately and integrates into the system. So surely, the AS/400 should have a relational database. Again, surveys have shown that many AS/400 customers planned to buy a database in the near future. I should note that you should always question survey results; they may say as much about the survey as they do about the AS/400 and its customers.

We at IBM also have had difficulties talking about the AS/400 database, because it is not something we can point to. The AS/400 database is not in only one place; again, the integrated nature of the system makes the database pervasive throughout the system. Conventional databases, on the other hand, are separate software components on top of an operating system. A program has to go through the operating system or use a totally separate interface to get to a conventional database.

By being integrated, partly above the MI and partly in SLIC, the AS/400 database achieves a higher level of efficiency than one built on top of an existing system. This database also is tightly integrated with other system components with which it communicates in a way no conventional database can achieve.