From: | "Tomas Vondra" <tv(at)fuzzy(dot)cz> |
---|---|
To: | "Gaëtan Allart" <gaetan(at)nexylan(dot)com> |
Cc: | "Tomas Vondra" <tv(at)fuzzy(dot)cz>, "pgsql-general(at)postgresql(dot)org" <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: General performance/load issue |
Date: | 2011-11-24 14:40:35 |
Message-ID: | f07635dcc0ce99df2a333c8ac6848b23.squirrel@sq.gransy.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On 24 Listopad 2011, 15:27, Gaëtan Allart wrote:
> Hi Thomas,
>
> I will be using iotop ;)
> Right now, most i/o come from "postgres: wal writer process".
What do you mean by "most I/O" - how much data is it writing? Is there a
vacuum running at the same time? What other processes are doing I/O?
Post a few lines of "iostat -x 5" so we know what kind of I/O we're
dealing with.
> -> dirty :
>
> cat /proc/sys/vm/dirty_ratio 20
> cat /proc/sys/vm/dirty_background_ratio 10
This means the cache may contain up to 3.2GB of data before the system
starts to write them out on background, and about 6.4GB before the
processes can't use the write cache.
What about /proc/sys/vm/dirty_expire_centiseconds?
Is there something interesting in the postgresql.log? Autovacuum maybe?
Tomas
From | Date | Subject | |
---|---|---|---|
Next Message | David Johnston | 2011-11-24 15:10:06 | Re: How to get normalized data from tekst column |
Previous Message | Gaëtan Allart | 2011-11-24 14:27:48 | Re: General performance/load issue |