Distribution information for tables in dbschema output

The dbschema -hd command with the name of the table retrieves the distribution information that is stored for a table in a database. If you specify the ALL keyword for the table name, the distributions for all the tables in the database are displayed.

During the dbimport operation, distribution information is created automatically for leading indexes on non-opaque columns. Run the UPDATE STATISTICS statement in MEDIUM or HIGH mode to create distribution information about tables that have the following types of indexes:
  • Virtual Index Interface (VII) or function indexes
  • Indexes on columns of user-defined data types
  • Indexes on columns of built-in opaque data types (such as BOOLEAN or LVARCHAR)
Output from the dbschema utility shows distribution information if you used the SAMPLING SIZE keywords when UPDATE STATISTICS in MEDIUM or HIGH mode ran on the table.

For information about using the UPDATE STATISTICS statement, see the HCL OneDB™ Guide to SQL: Syntax.

The output of dbschema for distributions is provided in the following parts:
  • Distribution description
  • Distribution information
  • Overflow information

Each section of dbschema output is explained in the following sections. As an example, the discussion uses the following distribution for the fictional table called invoices. This table contains 165 rows, including duplicates.

You can generate the output for this discussion with a call to dbschema that is similar to the following example: 
dbschema -hd invoices -d pubs_stores_demo