Solving insert performance

Insert comes under  data manipulation language. Insert means to place a new row into a table, which could be a duplicate or unique in itself.

The basic syntax of insert goes like this:

INSERT INTO table_name [(col_name [, ..., col_name])]
VALUES (expression [, ..., expression]);

A variation called INSERT SELECT takes data from an already populated table and inserts it into another one:

INSERT INTO table_name [(col_name [, ..., col_name])] SELECT statement;

Getting ready

To complete this recipe, you will need to connect to the Teradata database instance and open SQLA. 

How to do it...

Insert is the topmost SQL statement that executes on the system. With a little tweak, we can optimize it to get better performance:

1. Create a source table:

/**Source Table**/
CREATE SET TABLE TEST01.web_clicks
  (Web_page_Number INTEGER,   
   Location_Number INTEGER,
   Amount DECIMAL(10,2))
UNIQUE PRIMARY INDEX ( Web_page_Number );

2. Create a target table with the same PI as the source table:

...