Message-ID: <454F441B.1090601@alteeve.com>
Date: Mon, 06 Nov 2006 09:18:03 -0500
From: Madison Kelly <linux@alteeve.com>
User-Agent: Thunderbird 1.5.0.7 (X11/20060909)
MIME-Version: 1.0
To: Tobias Brox <tobias@nordicbet.com>
CC: Scott Marlowe <smarlowe@g2switchworks.com>,
	pgsql-performance@postgresql.org
Subject: Re: Setting "nice" values
References: <20061102160057.GA26580@oppetid.no> <454F3434.4030205@alteeve.com>
	<20061106133353.GA22050@oppetid.no> <454F3D23.6040003@alteeve.com>
	<20061106141149.GA7642@oppetid.no>
In-Reply-To: <20061106141149.GA7642@oppetid.no>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Tobias Brox wrote:
> [Madison Kelly - Mon at 08:48:19AM -0500]
>> Ah, sorry, long single queries is what you meant. 
> 
> No - long running single transactions :-)  If it's only read-only
> queries, one will probably benefit by having one transaction for every
> query.
> 

In this case, what happens is one kinda ugly big transaction is read 
into a hash, and then looped through (usually ~10,000 rows). On each 
loop another, slightly less ugly query is performed based on the first 
query's values now in the hash (these queries being where throttling 
might help). Then after the second query is parsed a PDF file is created 
(also a big source of slowness). It isn't entirely read-only though 
because as the PDFs are created a flag is updated in the given record's 
row. So yeah, need to experiment some. :)

Madi