Message-ID: <548A3103.70203@catalyst.net.nz>
Date: Fri, 12 Dec 2014 13:04:19 +1300
From: Mark Kirkwood <mark.kirkwood@catalyst.net.nz>
User-Agent: Mozilla/5.0 (X11; Linux x86_64;
 rv:31.0) Gecko/20100101 Thunderbird/31.3.0
MIME-Version: 1.0
To: Eric Pierce <epierce@saasmadeeasy.com>,
 Evgeniy Shishkin <itparanoia@gmail.com>,
 Andrea Suisani <sickpig@opinioni.net>
CC: "mfatticcioni@mbigroup.it" <mfatticcioni@mbigroup.it>,
 "pgsql-performance@postgresql.org" <pgsql-performance@postgresql.org>
Subject: Re: Tuning the configuration
References: <548823F8.4070003@mbigroup.it> <548987D7.7040304@opinioni.net>,
 <6B084BAC-82C6-410A-88D4-9CDC01DA8C09@gmail.com>
 <cf0230bc733d4f9aa7d7d0d35e2ffa4f@H-EX-01.hosting.local>
In-Reply-To: <cf0230bc733d4f9aa7d7d0d35e2ffa4f@H-EX-01.hosting.local>
Content-Type: text/plain; charset=windows-1252; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk
Sender: pgsql-performance-owner@postgresql.org

On 12/12/14 11:36, Eric Pierce wrote:
>
> ________________________________________
> From: pgsql-performance-owner@postgresql.org <pgsql-performance-owner@postgresql.org> on behalf of Evgeniy Shishkin <itparanoia@gmail.com>
> Sent: Thursday, December 11, 2014 7:11 AM
> To: Andrea Suisani
> Cc: mfatticcioni@mbigroup.it; pgsql-performance@postgresql.org
> Subject: Re: [PERFORM] Tuning the configuration
>
>> On 11 Dec 2014, at 15:02, Andrea Suisani <sickpig@opinioni.net> wrote:
>>
>> On 12/10/2014 11:44 AM, Maila Fatticcioni wrote:
>>> 2- I would like to use the two SDD to store the wal file. Do you think
>>> it is useful or how should I use them?
>>
>> I definitely would give it a try.
>>
>
>
>> I don't understand the logic behind using drives,
>> which are best for random io, for sequent io workloads.
>
>> Better use 10k sas with BBU raid for wal, money wise.
>
> Very much agree with this.  Because SSD is fast doesn't make it suited for certain things, and a streaming sequential 100% write workload is one of them.   I've worked with everything from local disk to high-end SAN and even at the high end we've always put any DB logs on spinning disk.  RAID1 is generally sufficient.  SSD is king for read heavy random I/O workload.
>


Mind you wal is a little different - the limiting factor is (usually) 
not raw sequential speed but fsync latency. These days a modern SSD has 
fsync response pretty much equal to that of a card with BBU + spinners - 
and has "more" high speed storage available (cards usually have only a 
1G or so of RAM on them).


regards

Mark


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