Re: Deterioration in performance when query executed in multi threads

Hi Thomas,
It is not a dedicated box (we have Jboss running too).

cpu_tuple_cost | 0.01
seq_page_cost | 1
random_page_cost | 4
effective_cache_size | 512MB

We have the data directory on nfs (rw,intr,hard,tcp,rsize=32768,wsize=32768,nfsvers=3,tcp). Note that we have also tested putting the data directory on local disk and didn't find a big improvement.

Thanks,
Anne

-----Original Message-----
From: pgsql-performance-owner(at)postgresql(dot)org [mailto:pgsql-performance-owner(at)postgresql(dot)org] On Behalf Of Thomas Kellerer
Sent: Monday, May 06, 2013 10:12 AM
To: pgsql-performance(at)postgresql(dot)org
Subject: Re: [PERFORM] Deterioration in performance when query executed in multi threads

Anne Rosset, 06.05.2013 19:00:
> Postgres version: 9.0.13
>
>> Work_mem is set to 64MB
>> Shared_buffer to 240MB
>> Segment_size is 1GB
>> Wal_buffer is 10MB
>
> Artifact table: 251831 rows
> Field_value table: 77378 rows
> Mntr_subscription: 929071 rows
> Relationship: 270478 row
> Folder: 280356 rows
> Item: 716465 rows
> Sfuser: 5733 rows
> Project: 1817 rows
>
> 8CPUs
> RAM: 8GB
>

With 8GB RAM you should be able to increase shared_buffer to 1GB or maybe even higher especially if this is a dedicated server.
240MB is pretty conservative for a server with that amount of RAM (unless you have many other applications running on that box)

Also what are the values for

cpu_tuple_cost
seq_page_cost
random_page_cost
effective_cache_size

What kind of harddisk is in the server? SSD? Regular ones (spinning disks)?

--
Sent via pgsql-performance mailing list (pgsql-performance(at)postgresql(dot)org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-performance