Some Terminology Used in DB2

The terminology used to describe the database objects has changed again in DB2 for Linux, UNIX, and Windows Version 8, and this can lead to confusion for those experienced with previous versions of the product. This section will focus on the current terms but will point out where some of the old terminology may still be used.

A database can be described in the simplest terms as a collection of data. If you are familiar with DB2 or any other Relational Database Management Systems (RDBMSs), you'll know that data within a database is stored in tables. Tables are defined with a set of columns and the data is stored in rows within these tables. In DB2 Enterprise Server Edition (ESE), a database can be divided into several parts that are called database partitions. A database partition is a part of the database that has its own portion of the user data, indexes, configuration files, and transaction logs.


To use more than one database partition, the Database Partitioning Feature must be acquired from IBM.

In the past, the term node was used instead of partition; however, this was very confusing because the term node was used to describe both the software and the hardware. A node in DB2 terminology is equivalent to a database partition, however, the computers involved in the database cluster were also referred to as nodes. This book will use the new terminology to avoid confusion, but a number of the DB2 commands and SQL statements also support the old terminology to support scripts written for previous versions of DB2.