Newbie--indexes

Can someone briefly explain basic indexing strategy? Indexes on single cols vs. multiple cols? Is there a good reference?

(What inspired this question is a query I was running today on a million-row table with an ad-hoc relationship to a short table I was trying to use to select rows from the long one.

Providing a literal key value in the query gave results in seconds; selecting by matching to the keys in the short file ran for 2 hours (on an HP-UX I-40) before I cancelled. Obviously, there is something I am not getting here . . .

Thanks. Received on Fri Jul 28 1995 - 00:00:00 CEST