Re: Postgres for a "data warehouse", 5-10 TB

From: Scott Marlowe <scott(dot)marlowe(at)gmail(dot)com>
To: Robert Klemme <shortcutter(at)googlemail(dot)com>
Cc: pgsql-performance(at)postgresql(dot)org
Subject: Re: Postgres for a "data warehouse", 5-10 TB
Date: 2011-09-12 19:08:04
Message-ID: CAOR=d=0U=+0v0UkpBacBTJh06rAwScpWMCvFrDO26OUA7jsctw@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

On Mon, Sep 12, 2011 at 11:04 AM, Robert Klemme
<shortcutter(at)googlemail(dot)com> wrote:
> On 11.09.2011 22:10, Scott Marlowe wrote:
>
>> Another data point.  We had a big Oracle installation at my last job,
>> and OLAP queries were killing it midday, so I built a simple
>> replication system to grab rows from the big iron Oracle SUN box and
>> shove into a single core P IV 2.xGHz machine with 4 120G SATA drives
>> in SW RAID-10.
>>
>> That machine handily beat the big iron Oracle machine at OLAP queries,
>> running in 20 minutes what was taking well over an hour for the big
>> Oracle machine to do, even during its (Oracle machine) off peak load
>> times.
>
> Um, that sounds as if the SUN setup was really bad.  Do you remember any
> details about the drive configuration there?

It was actually setup quite well. A very fast SAN with individual
drive arrays etc. It was VERY fast at transactional throughput. BUT
it was not setup for massive OLAP work. The drives that housed the
statistical data we were running OLAP against were the slowest in the
set, since they were made to mostly just take in a small amount of
data each minute from the java servers. Originally the stats had been
on a pg server in production and very fast, but some political
decision moved it onto the Oracle server. The Oracle DBA wasn't any
happier with this move than me, btw.

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message pasman pasmański 2011-09-12 19:12:02 Re: Allow sorts to use more available memory
Previous Message Andy Colson 2011-09-12 18:57:09 Re: Allow sorts to use more available memory