Discovery compiles reverse indexes of the data in the data sources to speed up discovery of and access to it. The indexes are in turn used for relationship discovery.

Accessing the actual data, either when building the index or when performing relationship discovery, requires that Discovery understand the diverse data file formats of all of the data sources. Discovery uses the JDBC standard to access databases.

Discovery uses high-performance federated query algorithms to optimize the select statements generated by the search engine for relationship discovery.

Indexing speed depends on network speed, database constraints, table size, and other traits of the data sources.

Discovery has configuration parameters with which you can control the indexing process. For example, you can configure Discovery to ignore case during indexing, or use data sampling for very large data sources.