Amazon Redshift
Database Developer Guide (API Version 2012-12-01)

Compute Node–Only Functions

Some Amazon Redshift queries must execute only on the compute nodes. If a query references a user-created table, the SQL runs on the compute nodes.

A query that references only catalog tables (tables with a PG prefix, such as PG_TABLE_DEF) or that does not reference any tables, runs exclusively on the leader node.

If a query that uses a compute-node function doesn't reference a user-defined table or Amazon Redshift system table returns the following error.

[Amazon](500310) Invalid operation: One or more of the used functions must be applied on at least one user created table.

The documentation for each compute-node only function includes a note stating that the function will return an error if the query doesn't references a user-defined table or Amazon Redshift system table.

The following SQL functions are compute-node only functions: