Which index to create ?

Hi,

  I have a question regarding the tuning of one of my SQL statements. Suppose there is a table X in my Schema which has columns A, B, C (all INTEGER) where A is the primary key. Suppose further that the table is pretty large: 10 million rows or more. Quite frequently I want to do this statement:

SELECT DISTINCT B FROM X The explain plan tells me there is a full table scan on X. So I thought it might help to create an index on the B column but I still get a full table scan. I believe that the full table scan is the main performance bottleneck... how can I get rid of the full table scan ?

André Received on Mon Dec 08 2003 - 02:21:24 CST