"Jonah H. Harris" <[EMAIL PROTECTED]> writes:
> On 3/7/06, Tom Lane <[EMAIL PROTECTED]> wrote:
> >
> > However, now that we've changed the code to prefer large numbers of tapes,
> > it's not at all clear that Algorithm D is still the right one to use. In
> > particular I'm looking at cascade merge, Algorithm 5.4.3C, which appears
> > to use significantly fewer passes when T is large. Do you want to try
> > that?
>
> Guess we won't really know 'til it can be tested :)
It would also be interesting to allow multiple temporary areas to be declared
and try to spread tape files across the temporary areas. Ideally keeping input
and output tapes on separate drives.
--
greg
---------------------------(end of broadcast)---------------------------
TIP 4: Have you searched our list archives?
http://archives.postgresql.org
