On Sun, 2004-06-20 at 17:15, Tatsuo Ishii wrote:
> > > Also I think we need to enhance ALTER INDEX to assign new table spaces
> > > for indexes. Assigning different tables spaces for tables and indexes
> > > are essential to gain more I/O speed IMO.
> >
> > I thought about this. ALTER INDEX doesn't exist yet and I figured that,
> > unlike the case of tables, its easy to drop and recreate indexes in new
> > tablespaces.
>
> Oh you are right. I forgot about CREATE INDEX ... TABLESPACE.
>
> > I'm still stumped as to where I am corrupting memory with this patch
> > though. (There was another bug: I wasn't detecting the case where users
> > set tablespace to the tablespace that the table is already in).
On a related note, will there be a way to have implicit index creation
occur in a seperate table space automagically? I.e.
create table test (id int4 primary key, n1 int unique);
so that the indexes created in id and n1 here would have a different
default namespace than the table? Just wondering.