Thread: Data partitioning?

hi,

I am becoming more and more convinced that in order to achieve the
required performance and scalability I need to split my data amoungst
many backend machines.

Ideally I would start with about 10 machine and have 1/10th of the data
on each. As the data set grows I would then buy additional machines to
spread the data over.

I think this is called data partitioning in oracle?
What does postgresql have to offer in this area?

Does software already exist to query multiple identical schemas each
with different data stored in them, and then correlate the results?

If not what suggestions does anyone have for writing my own?

Thanks,
Mat

--

On Tue, Oct 07, 2003 at 10:34:18AM +0100, psql-mail@freeuk.com wrote:
> I am becoming more and more convinced that in order to achieve the
> required performance and scalability I need to split my data amoungst
> many backend machines.

How much data do you have, and what are you trying to run on?  That
sounds like a _lot_.   In any case, you can't do this in PostgreSQL
right now.  A single back end controls all the data.

> I think this is called data partitioning in oracle?

Usually people mean spreading tables across different disks by some
criteria when they talk about that.  Completely different machines is
pretty usual.

> If not what suggestions does anyone have for writing my own?

Be very careful.  Multi-database concurrency problems are easy to
cause and hard to solve.

A

----
Andrew Sullivan                         204-4141 Yonge Street
Afilias Canada                        Toronto, Ontario Canada
<andrew@libertyrms.info>                              M2P 2A8
                                         +1 416 646 3304 x110