>Glenn,
>
>You say
>"...you probably can maintain an index on Deleted()because there are so few so as not to create a problem.".
>
>But that is precisely the problem - with only a small percentage of entries that *are* Deleted(), then given a reasonably large number of records performance will be adversely affected every time when running with DELETED set ON.
>
>It is NOW known that an index on DELETED() is a bad thing *most* of the time.
>
>Cheers,
>
>JimN

Hi Jim,

I am not advocating indexing on deleted, in fact by using the technique as described you only need the primary key to locate deleted records. It was my understanding however from benchmarking we performed that it is only when there are many repeating entries in a given key set that performance suffers. Do you have results that differ with my findings? How did you conduct the test?

Glenn