Microsoft® SQL Server™ 2008 All-In-One Desk Reference For Dummies®

VIII.6.2. Partitioning Key Terms and Concepts

Before getting started on a simple partitioning example, it's worth understanding some underlying architectural concepts and terminology.

• Partition type: An administrator might partition tables based on data or index values. Partitions based on data are range partitions. Database designers typically use an easy-to-understand factor when creating a range partition. For example, dates, regions, and other easily explained values are ideal for distributing information to multiple destinations. You might also associate indexes with your partition, which streamlines access to your data and related indexes. In fact, if you create an index on an already partitioned table, SQL Server automatically places the index into the appropriate partition.

• Partition key: This is a column that SQL Server uses as input to a partition function (described in a moment). The database administrator identifies this column and cites it when creating the partitioning scheme and associated table definition. When new rows are created, or existing rows are modified, the partition function examines the value contained in this column to decide where its associated row is placed.

• Partition functions: These specialized functions, defined by a SQL Server database administrator, serve as logical filters that help identify the appropriate destination for a given row.

  Writing ...