From: | Craig Ringer <craig(at)2ndQuadrant(dot)com> |
---|---|
To: | Evgeny Shishkin <itparanoia(at)gmail(dot)com> |
Cc: | Niels Kristian Schjødt <nielskristian(at)autouncle(dot)com>, Jeff Janes <jeff(dot)janes(at)gmail(dot)com>, "pgsql-performance(at)postgresql(dot)org" <pgsql-performance(at)postgresql(dot)org> |
Subject: | Re: Do I have a hardware or a software problem? |
Date: | 2012-12-12 01:03:14 |
Message-ID: | 50C7D7D2.4070605@2ndQuadrant.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-performance |
On 12/12/2012 06:44 AM, Evgeny Shishkin wrote:
>
> On Dec 12, 2012, at 2:41 AM, Niels Kristian Schjødt
> <nielskristian(at)autouncle(dot)com <mailto:nielskristian(at)autouncle(dot)com>> wrote:
>
>> Are you using a hardware based raid controller with them?
>>
> Yes, of course. Hardware raid with cache and bbu is a must. You can't
> get fast fsync without it.
Most SSDs should offer fairly fast fsync without a hardware RAID
controller, as they do write-back caching. The trick is to find ones
that do write-back caching safely, so you don't get severe data
corruption on power-loss.
A HW RAID controller is an absolute must for rotating magnetic media,
though.
> Also mdadm is a pain in the ass and is suitable only on amazon and
> other cloud shit.
I've personally been pretty happy with mdadm. I find the array
portability it offers very useful, so I don't need to buy a second RAID
controller just in case my main controller dies and I need a compatible
one to get the array running again. If you don't need a BBU for safe
write-back caching then mdadm has advantages over hardware RAID.
I'll certainly use mdadm over onboard fakeraid solutions or low-end
hardware RAID controllers. I suspect a mid- to high end HW RAID unit
will generally win.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
From | Date | Subject | |
---|---|---|---|
Next Message | Evgeny Shishkin | 2012-12-12 01:17:14 | Re: Do I have a hardware or a software problem? |
Previous Message | Tom Lane | 2012-12-12 00:38:31 | Re: Why is PostgreSQL 9.2 slower than 9.1 in my tests? |