Re: Rapidly decaying performance repopulating a large table

From: Tomasz Ostrowski <tometzky(at)batory(dot)org(dot)pl>
To: David Wilson <david(dot)t(dot)wilson(at)gmail(dot)com>, pgsql-general(at)postgresql(dot)org
Subject: Re: Rapidly decaying performance repopulating a large table
Date: 2008-04-23 11:10:54
Message-ID: 480F193E.7090407@batory.org.pl
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

On 2008-04-22 23:46, David Wilson wrote:

> Upping the segments to 50, timeout to 30m and completion target to
> 0.9 has improved average copy time to between 2 and 10 seconds, which
> is definitely an improvement.

I'd up them to 128 (or even 256) and set completion target back to 0.5.
But make sure you'll always have 4GB (8GB) of disk space for wal logs.

All you'd risk is several minutes of recovering in case of server crash.

Regards
Tometzky
--
...although Eating Honey was a very good thing to do, there was a
moment just before you began to eat it which was better than when you
were...
Winnie the Pooh

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Terry Lee Tucker 2008-04-23 11:14:16 Re: Deny creation of tables for a user
Previous Message Pascal Cohen 2008-04-23 10:46:44 Deny creation of tables for a user