Determining the cost of disk space when creating an index (Simple and Compound) in eDirectory

Creating indexes in eDirectory greatly increases the speed with which commonly searched attribute values can be returned via LDAP searches.  However, this benefit comes with a cost.  There is the overhead of maintaining the index, the additional memory consumed as well as an increase in overall dib size.

It is the later that many customers first notice and is the focus of this TID.

Indexes, whether system or user created, move the attribute values that are to be indexed from the entry's container to its own attribute container with a link back to the entry. 

Below are some sample numbers from dibs ranging from 24,000 to 1 million objects and the additional disk space required for both simple and compound indexes using a value index.  It is our hope that these numbers from our own testing can be extrapolated to larger environments to help customers with their resource planning.