Re: Server cache for certain reference tables...

Peter's suggestion is fine, but as a design issue, consider instead of having many reference tables with seeing if you can have only a few (ideally one!).

The situation usually is to have many tables which basically only have two columns CODE and NAME. You can join these into a single table with columns CODE, NAME, CODE_TYPE where CODE_TYPE indicates what would have been the table name. Since these tables typically have few values, they al fit into a few Oracle blocks. This is breaking at least one of Codd's rules, and I really don't know or care which, but is a practical approach I've seen used to good effect in several projects.

Dave O'Keeffe

Peter Rak wrote:
>
> Hi,
> try alter table table_name cache;
>
> Peter.
>
> kolli_s_at_hotmail.com wrote:
>
> > Hi,
> >
> > According to our database model we will end up joining the main data
> > table with several smaller reference tables for almost every select
> > query. I am trying to find if there is a way in Oracle to keep certain
> > tables in buffer (cache or whatever) *all the time* so that our joins
> > do not take too long. I know that it's possible in Sybase (11.x.x).
> >
> > -----== Posted via Deja News, The Leader in Internet Discussion ==-----
> > http://www.dejanews.com/ Now offering spam-free web-based newsreading

--
For email replys remove "n0spam." from the above address Received on Tue Apr 21 1998 - 15:18:58 CDT