This chapter describes Teradata Database primary indexes: unique versus nonunique and row partitioned versus nonpartitioned.
The process of selecting a primary index is given heavy emphasis. Other topics describe primary index access to rows, the various performance considerations that pertain to primary indexes, special considerations for nonunique primary indexes relating to duplicate row checking, and space utilization parameters for primary indexes.
The main emphasis of the chapter is primary index features dealing with user data tables, but the considerations for selecting the column set to be used for the primary index for a base data table are largely identical to those for selecting the column set to be used as the primary index for a join index. Any join index‑specific considerations for primary indexes are documented in Chapter 11: “Join and Hash Indexes.”
The chapter also covers tables and join indexes that do not have a primary index. These database objects are referred to as NoPI objects.
NoPI objects come in two forms.