Hi Seth, Juergen:
I have similar problem here with a raid0 using raidtools
#mkfs -t xfs -f -l version=2 /dev/md1
meta-data=/dev/md1 isize=256 agcount=29, agsize=1048544
blks
= sectsz=512
data = bsize=4096 blocks=29902912, imaxpct=0
= sunit=32 swidth=64 blks, unwritten=0
naming =version 2 bsize=4096
log =internal log bsize=4096 blocks=3680, version=2
= sectsz=512 sunit=32 blks
realtime =none extsz=262144 blocks=0, rtextents=0
# mount -t xfs -o logbsize=64k /dev/md1 /mnt
mount: wrong fs type, bad option, bad superblock on /dev/md1,
or too many mounted file systems
#dmesg | tail -1
XFS: logbuf size must be greater than or equal to log stripe size
Info
----
OS: RH 7.3 installed with fixed RH7.3-SGI-XFS-1.1.iso
(new-rh-7.3-patch.xdelta)
kernel-2.4.19-SGI_XFS_1.2.0smp
xfsprogs-2.3.9
mount-2.11n-12.7.3
Before I tried to use mkfs with this options:
mkfs -t xfs -l size=32768b,version=2 -f /dev/md1
and tried to mount with:
mount -t xfs -o logbufs=8,logbsize=32768 /dev/md1 /mnt and I get the
same error.
Any help will be appreciate.
Thanks
Benito.-
On Thu, 26 Jun 2003, Seth Mos wrote:
> At 13:38 26-6-2003 +0200, Juergen Rose wrote:
> >Hi,
> >
> >I have here a software raid device /dev/md/0. If I create the filesystem
> >with:
> > mkfs -t xfs -f /dev/md/0
> >I can mount the device, but I get a lot of "raid5: switching cache
> >buffer size" warnings. If I create the filesystem with
> > mkfs -t xrs -f -l version=2 /dev/md/0
> >I can't mount the device: I get:
> >
> >mount: wrong fs type, bad option, bad superblock on /dev/md/0,
> > or too many mounted file systems
>
> It probably needs to be mounted with a larger buffer size.
>
> Try mount -o logbsize=64k /dev/foo /mnt/foo
>
> The version2 log automaticaly switches to a larger databuffer size for
> striping purposes.
> However if it is larger then 32k a mount option must be specified.
>
> Cheers
>
> --
> Seth
> It might just be your lucky day, if you only knew.
>
>
--
|