Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 28 Aug 2010 13:05:49 -0300
From:      Thiago Damas <tdamas@gmail.com>
To:        Ilya Bakulin <webmaster@kibab.com>
Cc:        =?ISO-8859-1?Q?Dag=2DErling_Sm=F8rgrav?= <des@des.no>, hackers@freebsd.org, geom@freebsd.org
Subject:   Re: Support for WD Advanced Format disks
Message-ID:  <AANLkTikwgiw7r1KFdMVYaoomMGmvpjw=2rGtqM%2BZ3at2@mail.gmail.com>
In-Reply-To: <20100828195024.3d671a76@kibab.com>
References:  <86wrry1hwv.fsf@ds4.des.no> <20100828130912.48205a47@kibab.com> <20100828195024.3d671a76@kibab.com>

next in thread | previous in thread | raw e-mail | index | archive | help
You read:
"ATA 4K sector issues"
http://lists.freebsd.org/pipermail/freebsd-hackers/2010-March/031154.html

2010/8/28 Ilya Bakulin <webmaster@kibab.com>:
> On Sat, 28 Aug 2010 13:09:12 +0400
> Ilya Bakulin <webmaster@kibab.com> wrote:
>
>> Right now I'm running phybs -w /dev/ad7p1
>
> Well, phybs finished.
> Results:
> kibab-desktop# ./phybs -w /dev/ad7p1
> =A0 count =A0 =A0size =A0offset =A0 =A0step =A0 =A0 =A0 =A0msec =A0 =A0 t=
ps =A0 =A0kBps
>
> =A0131072 =A0 =A01024 =A0 =A0 =A0 0 =A0 =A04096 =A0 =A0 3098125 =A0 =A0 =
=A0 0 =A0 =A0 =A042
> =A0131072 =A0 =A01024 =A0 =A0 512 =A0 =A04096 =A0 =A0 3100047 =A0 =A0 =A0=
 0 =A0 =A0 =A042
>
> =A0 65536 =A0 =A02048 =A0 =A0 =A0 0 =A0 =A08192 =A0 =A0 1505809 =A0 =A0 =
=A0 0 =A0 =A0 =A087
> =A0 65536 =A0 =A02048 =A0 =A0 512 =A0 =A08192 =A0 =A0 1505179 =A0 =A0 =A0=
 0 =A0 =A0 =A087
> =A0 65536 =A0 =A02048 =A0 =A01024 =A0 =A08192 =A0 =A0 1506127 =A0 =A0 =A0=
 0 =A0 =A0 =A087
>
> =A0 32768 =A0 =A04096 =A0 =A0 =A0 0 =A0 16384 =A0 =A0 1522343 =A0 =A0 =A0=
 1 =A0 =A0 =A086
> =A0 32768 =A0 =A04096 =A0 =A0 512 =A0 16384 =A0 =A0 1521017 =A0 =A0 =A0 1=
 =A0 =A0 =A086
> =A0 32768 =A0 =A04096 =A0 =A01024 =A0 16384 =A0 =A0 1523748 =A0 =A0 =A0 1=
 =A0 =A0 =A086
> =A0 32768 =A0 =A04096 =A0 =A02048 =A0 16384 =A0 =A0 1522853 =A0 =A0 =A0 1=
 =A0 =A0 =A086
>
> =A0 16384 =A0 =A08192 =A0 =A0 =A0 0 =A0 32768 =A0 =A0 =A0890911 =A0 =A0 =
=A0 3 =A0 =A0 147
> =A0 16384 =A0 =A08192 =A0 =A0 512 =A0 32768 =A0 =A0 =A0892026 =A0 =A0 =A0=
 3 =A0 =A0 146
> =A0 16384 =A0 =A08192 =A0 =A01024 =A0 32768 =A0 =A0 =A0890699 =A0 =A0 =A0=
 3 =A0 =A0 147
> =A0 16384 =A0 =A08192 =A0 =A02048 =A0 32768 =A0 =A0 =A0891862 =A0 =A0 =A0=
 3 =A0 =A0 146
> =A0 16384 =A0 =A08192 =A0 =A04096 =A0 32768 =A0 =A0 =A0889903 =A0 =A0 =A0=
 3 =A0 =A0 147
>
> Obviously 34-sectors offset is bad for phybs. Write requests are NEVER ge=
t aligned.
> I have recreated partition as follows:
>
> kibab-desktop# gpart delete -i 1 ad7
> ad7p1 deleted
> kibab-desktop# gpart add -t freebsd-ufs -s 10G -b 63 ad7
> ad7p1 added
> kibab-desktop# gpart show ad7
> =3D> =A0 =A0 =A0 =A034 =A02930277101 =A0ad7 =A0GPT =A0(1.4T)
> =A0 =A0 =A0 =A0 =A034 =A0 =A0 =A0 =A0 =A029 =A0 =A0 =A0 - free - =A0(15K)
> =A0 =A0 =A0 =A0 =A063 =A0 =A020971520 =A0 =A01 =A0freebsd-ufs =A0(10G)
> =A0 =A020971583 =A02909305552 =A0 =A0 =A0 - free - =A0(1.4T)
>
> So, we have ad7p1 which is still misaligned, but if write request will st=
art at offset=3D512, it will be aligned properly.
> Let's test as follows:
>
> kibab-desktop# ./phybs -l 4096 -w /dev/ad7p1
> =A0 count =A0 =A0size =A0offset =A0 =A0step =A0 =A0 =A0 =A0msec =A0 =A0 t=
ps =A0 =A0kBps
>
> =A0 32768 =A0 =A04096 =A0 =A0 =A0 0 =A0 16384 =A0 =A0 1520609 =A0 =A0 =A0=
 1 =A0 =A0 =A086
> =A0 32768 =A0 =A04096 =A0 =A0 512 =A0 16384 =A0 =A0 =A0 17898 =A0 =A0 151=
 =A0 =A07323
> =A0 32768 =A0 =A04096 =A0 =A01024 =A0 16384 =A0 =A0 1524899 =A0 =A0 =A0 1=
 =A0 =A0 =A085
>
> :-) Aligned writes are 86 times faster than misaligned!!!
> --
> Regards,
> Ilya Bakulin
> http://kibab.com
> xmpp://kibab612@jabber.ru
>



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?AANLkTikwgiw7r1KFdMVYaoomMGmvpjw=2rGtqM%2BZ3at2>