Vantage hashes the primary index value set to determine on which AMPS the rows of a primary-indexed table are to be stored. For information about how the rows of tables and join indexes that do not have a primary index are distributed to the AMPs, see Teradata Vantage™ - Database Design, B035-1094.
Selection and update operations that use the primary index column in a WHERE clause use hashing to determine the location of any row with maximum efficiency.
You cannot compress the values of columns that are members of the primary index column set, nor can you compress the values of the columns of a partitioning expression.
You cannot include BLOB, CLOB, JSON, XML, BLOB-based UDT, CLOB-based UDT, XML-based UDT, Geospatial, ARRAY, VARRAY, Period, or row-level security constraint columns in the primary index column set, nor can you include row-level security constraint columns. Any table that contains BLOB, CLOB, Geospatial, ARRAY, VARRAY, Period, or row-level security constraint columns must contain at least 1 other non-BLOB, non-CLOB, non-Geospatial, non-ARRAY/VARRAY, or non-Period column that can be used to define its primary index.