Clustered Index / Non-Clustered Index


A clustered index (SQL Server, MySQL/InnoDB) is a table stored in an index B-Tree structure. There is no second data structure (heap-table) for the table.

A non-clustered index is an index that refers to another data structure containing further table columns.

Accessing table data via a secondary index (index on a clustered index) is slower than a similar query on a heap-table.

SQL Server supports clustered index optionally. You have a free choice between clustered indexes and heap-tables. There can be at most one clustered index per table. Dropping a clustered index transforms the table into a heap-table. Adding a clustered index to a heap table actually drops the heap structure. SQL Server supports non-unique clustered indexes on arbitrary columns. Creating an SQL Server table without clustering index requires the use of the NONCLUSTERED clause:

CREATE TABLE (
   id    NUMBER NOT NULL,
   [...]
   CONSTRAINT pk PRIMARY KEY NONCLUSTERED (id)
)

The MySQL InnoDB engine has mandatory clustered indexes. That means there is always a clustered index, often using the primary key. If there is no suitable unique key available, MySQL will use a generated row ID for that purpose. The MyISAM storage engine doesn’t support clustered indexes and uses heap-tables all the time.

The Oracle database has optional clustered indexes called Index-Organized Tables. They work on the primary key only.

About the Author

Photo of Markus Winand
Markus Winand tunes developers for high SQL performance. He also published the book SQL Performance Explained and offers in-house training as well as remote coaching at http://winand.at/