Message-ID: <5d7d4ed635ce16023d8871f692053242.squirrel@sq.gransy.com>
In-Reply-To: <53D862A0.80807@agliodbs.com>
References: <53D862A0.80807@agliodbs.com>
Date: Wed, 30 Jul 2014 11:01:55 +0200
Subject: Re: Why you should turn on Checksums with SSDs
From: "Tomas Vondra" <tv@fuzzy.cz>
To: "Josh Berkus" <josh@agliodbs.com>
Cc: pgsql-performance@postgresql.org
User-Agent: SquirrelMail/1.4.22
MIME-Version: 1.0
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: 8bit
Importance: Normal
Precedence: bulk
Sender: pgsql-performance-owner@postgresql.org

On 30 Červenec 2014, 5:12, Josh Berkus wrote:
> Explained here:
> https://www.usenix.org/system/files/conference/fast13/fast13-final80.pdf
>
> 13 out of 15 tested SSD's had various kinds of corruption on a power-out.
>
> (thanks, Neil!)

Well, only four of the devices supposedly had a power-loss protection
(battery, capacitor, ...) so I guess it's not really that surprising the
remaining 11 devices failed in a test like this. Although it really
shouldn't damage the device, as apparently happened during the tests.

Too bad they haven't mentioned which SSDs they've been testing
specifically. While I understand the reason for that (HP Labs can't just
point at products from other companies), it significantly limits the
usefulness of the study. Too many companies are producing crappy
consumer-level devices, advertising them as "enterprise". I could name a
few ...

Maybe it could be deciphered using the information in the paper
(power-loss protection, year of release, ...).

I'd expect to see Intel 320/710 to see there, but that seems not to be the
case, because those devices were released in 2011 and all the four devices
with power-loss protection have year=2012. Or maybe it's the year when
that particular device was manufactured?


regards
Tomas


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