Message-ID: <4EC46D46.5050608@squeakycode.net>
Date: Wed, 16 Nov 2011 20:11:18 -0600
From: Andy Colson <andy@squeakycode.net>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.9.2.21) Gecko/20110903 Lightning/1.0b2 Thunderbird/3.1.13
MIME-Version: 1.0
To: Tory M Blue <tmblue@gmail.com>
CC: pgsql-performance@postgresql.org
Subject: Re: Performance question 83 GB Table 150 million rows,
 distinct select
References: 
 <CAEaSS0b7D4eyDi4kt-Dt8u-yGq+akD=5PZwMk_TUHP-zujAnCQ@mail.gmail.com>
In-Reply-To: 
 <CAEaSS0b7D4eyDi4kt-Dt8u-yGq+akD=5PZwMk_TUHP-zujAnCQ@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

On 11/16/2011 04:53 PM, Tory M Blue wrote:
> Linux F12 64bit
> Postgres 8.4.4
> 16 proc / 32GB
> 8 disk 15KRPM SAS/Raid 5 (I know!)
>
>
> shared_buffers = 6000MB			
> #temp_buffers = 8MB			
> max_prepared_transactions = 0		
> work_mem = 250MB				
> maintenance_work_mem = 1000MB		
> 	
>
>
>
>
> We now have about 180mill records in that table. The database size is
> about 580GB and the userstats table which is the biggest one and the
> one we query the most is 83GB.
>
> Just a basic query takes 4 minutes:
>
> For e.g. select count(distinct uid) from userstats where log_date>'11/7/2011'
>

How'd you feel about keeping a monthly summary table?  Update it daily, with only a days worth of stats, then you could query the summary table much faster.

That's what I do for my website stats.  I log details for a month, then summarize everything into a summary table, and blow away the details.  You wouldn't have to delete the details if you wanted them, just keeping the summary table updated would be enough.

-Andy