Software Architect's Handbook

Primary/clustered indexes

When designing a table, one approach is to keep the rows unordered and create as many secondary indexes as necessary. Such an unordered structure is known as a heap. Another approach is to create a primary index, also known as a clustered index, to order the rows by the primary key. It is common to have a clustered index on a table which is known as a clustered table. The only time you might not want one is if the table is very small (and one that you know will remain small over time), such that the overhead of storing and maintaining the index is not worth it when compared to simply searching the table.

The one or more columns that make up the primary key of a table are the columns that make up the index definition. ...