Re: Performance Issue

On Fri, 04 Sep 1998 03:03:33 GMT, sooz_at_pobox.com (Susan) wrote:

>Unique indexes ultimately take longer when massive updates are being
>done, because they need to check the entire index for a non-match
>before the record can be commited, else they error. On extremely
>large tables, that might make a difference.
>

<snip>

Sorry, but Oracle indexes are B-trees, so the insert algorithm can't help detecting a duplicate, since when it arrives at the insertion point for the new value, it's either already there (give up) or not (put it in). Non-unique indexes are exactly the same, except that there's never any point in giving up, and the new value is always inserted.

Eric. Received on Sat Sep 12 1998 - 00:00:00 CEST