Re: Oracle Myths

In article <f369a0eb.0205230725.7d6f5d7f_at_posting.google.com>, you said (and I quote):
> FFS was not available when I worked with PeopleSoft. But what I see from other
> applications is that performance for aggregations on columns pulled by FFS is
> about the same however you order the columns in the index. My thought is that
> it's because result sets from FFS are not sorted.
>

Agreed. It doesn't seem to matter much. I should have explained myself better. Hope the data I sent to Niall makes it clearer.

The big difference I've noticed is that FFS with a reverse order of columns in index (most selective first) is considerably slower than a range scan with least selective column first. For any access that involves a range scan, including aggregates.

For any access that involves only a single row or no range scan or aggregate at all, then there is no practical difference.

To me, this defeats the purpose of using reverse order column (most selective first) indexes.

They work same as any normal index for single row access.

For range scans or aggregates, they work significantly worse than least selective column first.

What would then be the point of using one of these?

-- 
Cheers
Nuno Souto
nsouto_at_optushome.com.au.nospam