On Tue, 2007-09-04 at 07:09 +0300, Mikko Partio wrote:
>
>
> On 9/4/07, Ow Mun Heng <Ow.Mun.Heng@xxxxxxx> wrote:
> Hi,
>
> I'm running out of space on one of my partitions and I still
> have not
> gotten all the data loaded yet. I've read that one could
> symlink the
> pg_pg_xlog directory to another drive. I'm wondering if I can
> do the
> same for specific tables as well.
>
>
> Create another tablespace to the new location and the ALTER TABLE ...
> TABLESPACE newtablespace.
>
OOooohhh... I didn't know one could use tablespaces like that. (I mean,
I did read the docs, but it just didn't register that it _can_ do
something like that)
additional question.. do I need to change the search_path?
>
> Thanks.
>
> I've already done a pg_dump of the entire schema but have not
> dropped /
> re-init the DB to another location cos I'm afraid I'll lose
> some items.
> (I've to drop the DB, format the partition, merge it w/
> another
> partition and re-init the DB then restore the DB from the
> dump)
>
> sigh.. wish it was easier, (meaning, like SQL Server where one
> can
> detach an entire DB/tablespace and then re-attach it
> elsewhere)
>
> If you are moving the whole cluster and can afford the downtime, you
> can shutdown the postmaster, move $PGDATA to a new location and then
> start postmaster from that new location.
It's not a cluster. Its a devel DB in my laptop so.. no issues w/
dropping everything and re-creating. Just exploring my options..
The tablespace thing looks/sounds interesting though...
>
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
