Re: Response time vs UNIQUE and DUPLICATE Indexes

I'd like to put money on there being some duplicated values in those unique columns. I assume that defining the indexes as unique will help the optimiser decide how to access the data. I can't believe that making them unique would have any overhead at all. Suggest you suck and see. I'd be interested to know the outcome. Mike. In article <57vpik$h7d@cssun.mathcs.emory.edu>, Appala.Naidu@KSNCABP03.S NSGATE.sprint.com writes > > Hi all, > > I have two tables which contain large no of records and > the keys have not been defined as UNIQUE, even though they contain > only unique records. The tables involve a fair amount of insertions > and updates. > > Will it help in reducing the processing time, if I make the keys > as unique? I was thinking whatever advantage we may get in update > because of unique key, will be nullified by insertion since > there is no need to check for duplicate. Will someone clarify? > > Please send me a reply to: > > appala.naidu@nsc.sprint.com > > Thanks in advance. -- Mik Hobbs