Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
Home -> Community -> Usenet -> c.d.o.server -> Re: Query Optimiser Question
Wow thats some answer you gave me there buddy :) I think I get the drift of
it though i.e. the optimiser does just what its meant to do and selects what
it thinks is the best solution rather than a number of best solutions or all
the solutions. However it would only be situationally the best solution as
the DB environment changes all the time (Yes?). Thus there is need for
recompilation of DBs statistics and then reassessment and calculation by the
optimiser of what it the best solution based on those statistics. If this
is a regular occurence then am i correct in assuming it would be a tremdous
overhead for systems that are very active 24 hours a day 7 days a week?
Thanks
Paul
Jaap W. van Dijk <j.w.vandijk_at_kpn.com> wrote in message
news:01bfc65a$c4c4a080$2d2b15ac_at_HKTGN0003049994...
> I think soultion space is a algebraical notion: if you regard all the
> permissible intervals of all the relevant parameters as pieces of axes in
a
> mutidimensional space, you've got the solution space: the space of all
> possible solutions. If you then define criteria specifying how the
> parameters are allowed to depend on each other, you limit yourself to a
> part of the total solution space. The criteria an optimizer uses
determines
> how big this reduction of the solution space will be.
>
> Jaap.
>
>
> Paul <paul_at_yeahright.com> schreef in artikel
> <Lk9X4.112$5Y.2956_at_news11-gui.server.ntli.net>...
> > I have a question sitting here infront of me :
> >
> > Describe how cost based query optimisers work and include an explanation
as
> > to how they reduce the solution space and what aspects are compared by
the
> > optimiser.
> >
> > Its just the solution space i am abit puzzled by. Does this refer to
the
> > amount of space taken up by solution strategies in the DBMS? If so how
does
> > taking up large amounts of so called solution space lead to problems?
> >
> > Thanx
> > Paul
> >
> >
> >
Received on Thu May 25 2000 - 00:00:00 CDT