Thanks.
I had a bit of luck with searching around.
I think if I create a TABLESPACE and point that to any disk partition or mount point
CREATE TABLESPACE tablespace_name
OWNER user_name
LOCATION directory_path;
then when creating a database or even a table I just need to include the TABLESPACE name .
CREATE DATABASE logistics
TABLESPACE ts_primary;
This will allow me to use the secondary disk for purely databases.
Thanks
On Tue, 19 Oct 2021, 06:51 Holger Jakobs, <holger@xxxxxxxxxx> wrote:
Am 18.10.21 um 19:04 schrieb Zahid Rahman:
> Hi
>
> I have two disks attached to my Raspberry pi 4 .
>
> One disk is a 2TB with a partition called postgres on /dev/sdb/sd1 .
> When creating databases this is where I wish to create them.
>
> This is not the disk with the operating system installed .
> If I install postgres with pgadmin using apt install,
> any database I create will be installed on disk /dev/sda/sda1 by default.
>
> Can you please give me a HOW-TO on choosing the partition on a
> particular disk when creating a database or perhaps even a table.
>
> Thanks
> Zahid
>
>
Hi Zahid,
when setting up a database cluster with initdb, you determine the
directory path for the cluster.
So make sure the directory path resides on the disk you want it to be.
What is also possible is using a symbolic link to redirect to another
directory on the intended disk.
On the other hand, it's not possible to have some databases of a cluster
on one drive and some on another. A database cluster cannot be
distributed among disks.
Since you can have any number of database clusters on a machine, this
doesn't impose a problem.
Regards,
Holger
--
Holger Jakobs, Bergisch Gladbach, Tel. +49-178-9759012
