Subject: Re: Typo in the RaidFrame guide.
To: David Laight <firstname.lastname@example.org>
From: Marcin Jessa <email@example.com>
Date: 05/05/2006 10:50:48
On Fri, 5 May 2006 08:04:30 +0100
David Laight <firstname.lastname@example.org> wrote:
> On Wed, May 03, 2006 at 08:16:27PM +0000, Marcin Jessa wrote:
> > One more thing. Maybe newfs -O 1 /dev/raid0a in the 15.3.6. Setting
> > up Filesystems should also mention -O 2 since UFS2 is suitable for
> > drives with more than 1 Terabyte capacity and it is already today
> > possible to get 750G SATA drives.
> Is it? The 'lazy' initialisation of inodes makes newfs quicker.
> But the inodes themselves are twice the size and will take up more
> system memory....
> FFSv1 should be able to handle filesystems with up to 2^32 fragments,
> With 8k fragments I make that 32TB.
> Since NetBSD currently has problems describing disks over 2TB, and
> there are known FS corruption issues over that limit, there is no
> necessity to use FFSv2.
> Remember the FS code is identical, only the sizes of some fields
So you're saying using FFSv2 over FFSv1 would add extra overhead and is
actually a bad thing to do?
Ok, in that case maybe newfs(8) should be changed.
Based on the info in it I chose FFSv2 instead of 1 for my system.
Select the filesystem-format
0 4.3BSD; This option is primarily used to
build root file systems that can be understood by older
1 FFS; normal fast-filesystem (default).
2 UFS2; enhanced fast-filesystem (suited for more than 1
Terabyte capacity, access control lists).