>We have a table with 95,000 records, indexed on an int8 type
(ClientNumber).
>It is currently taking up to 10 seconds to retrieve one row based upon the
>following statement:
>
>SELECT * FROM ClientTable WHERE ClientNumber=1234566789;
>
>The client numbers are random and there is currently no locking on the
>database. I am running SCO OpenServer, PostgreSQL version 6.5.3. Is there
>any way we can reduce the time taken by this query?
I have tried an EXPLAIN of the above statement, and it tells me that
Postgres is doing a Seq Scan on the table to return the results, even though
it should be using the index. How can I make sure that it will always use
the index when I specify ClientNumber?