Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 28 Nov 2008 18:07:35 +0100 (CET)
From:      Wojciech Puchar <wojtek@wojtek.tensor.gdynia.pl>
To:        Valentin Bud <valentin.bud@gmail.com>
Cc:        freebsd-questions <freebsd-questions@freebsd.org>
Subject:   Re: 5 TB server
Message-ID:  <20081128180448.Y5415@wojtek.tensor.gdynia.pl>
In-Reply-To: <139b44430811280640g69d4843bq276b9aa2c8aa725d@mail.gmail.com>
References:  <139b44430811280548x36915301i766bfb15f162c8ca@mail.gmail.com>  <20081128145705.A5057@wojtek.tensor.gdynia.pl> <139b44430811280640g69d4843bq276b9aa2c8aa725d@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
> tests with ZFS and
> some with UFS. On the ZFS side the most attractive thing is the backup
> and of course
> the easiness of administration.

don't get fooled.

>>
>> if i were you i would get any motherboard with 8 SATA ports, up to 8 1TB
>> disks, cheapest available CPU, good PCIe gigabit cards or two.
>
> 8x1TB with some mirroring + striping would equal how much in terms of
> available space? Sorry i have to do my homework regarding RAID :|.

gstripe of 4 gmirrors will give 4TB
graid3 or graid5 will give 7TB, but much lower I/O performance (same with 
ZFS raidz). if you mainly read/write large files it's OK.

With graid5 read performance is near gstripe only writes are slow, with 
graid3 or ZFS raidz both are slow.

To be exact - single reads are not slow, but I/O capacity is low so under 
concurrent reads it will be very slow.





Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20081128180448.Y5415>