From: | Ondrej Ivanič <ondrej(dot)ivanic(at)gmail(dot)com> |
---|---|
To: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: Table Design question for gurus (without going to "NoSQL")... |
Date: | 2011-11-21 02:08:57 |
Message-ID: | CAM6mieK=GNkm2F39Wymqc+LKa6isY27kq3Us18qEwRde=GuPEg@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Hi,
On 21 November 2011 00:33, Phoenix Kiula <phoenix(dot)kiula(at)gmail(dot)com> wrote:
> Hi.
>
> Want to start another thread, loosely related to the performance
> problems thread I have going.
>
> Need some DB design guidance from the gurus here.
>
> My big table now has about 70 million rows, with the following columns:
You can put different tables/indexes on different disks using
tablespaces. For example, one tablespace for all tables and another
one for all indexes (and change random_page_cost and seq_page_cost
appropriately ie. lower random_page_cost). It is a good idea to put
pg_xlog on the separate drive too.
Sometimes Postgres just can't utilise all available resources properly
and you can get results faster by running query over multiple
connections. It could be worth to investigate pg-poolII's parallel
query mode. You don't need multiple servers - just setup multiple PG
instances on the same physical machine (up to one PG instance per
core, with lower shared_mem, ...). Alternative could be parallel DB
like Postgres-XC (http://wiki.postgresql.org/wiki/Postgres-XC) ...
--
Ondrej Ivanic
(ondrej(dot)ivanic(at)gmail(dot)com)
From | Date | Subject | |
---|---|---|---|
Next Message | Lee Hachadoorian | 2011-11-21 02:14:14 | Re: Significant Digits in Floating Point Datatype |
Previous Message | Tomas Vondra | 2011-11-21 01:56:54 | Re: Installed. Now what? |