🚀 go-pugleaf

Oh great guru!  This humble one begs you to enlighten him with your
wisdom on the
matter of optimizing SQL scripts.

1) When doing a select/where/groupby or select/where/orderby, am I
correct in assuming
that the limitting operation (the where clause) is imposed before the
order or group
operation?  IE: When working with a _very_ large table, is there any
reason to separate
the operation into a select/into#temp/where followed by a
select/from#temp/groupby,
in order to avoid applying the orderby or groupby to the entire
orginal table?

2) I am writing a procedure that uses a series of select statements
and a number of temp
tables.  Is it more efficient to drop each temp table as soon as I am
finished with it
in order to free up the cache space, or am I better off avoiding the
drop operations
and just letting SQL drop them automatically all at once when the
procedure ends?

I am running SQL Server 6.5 under NT 4.0.


I would appreciate it if replies were cc'd to my email account:
brent@ugcs.caltech.edu

Thanks

Brent

--

--
Brent Eubanks					    brent@ugcs.caltech.edu