Hi,
i've got a problem on our new SMP machine with IDE-Raid and kernel 2.4.9
that's giving me a terrible headache.
Our configuration:
2 WD drives in the raid1 array
1 Maxtor drive on normal ide
mystery #1:
I can mount a ext2 filesystem twice without any errors. The two filesystems
contain different data and changes to the second mount don't affect the
first mount. Strange!
Here is a snippet from console with /dev/sda5 mounted twice:
/dev/sda5 / ext2 rw,usrquota,grpquota 0 0
none /proc proc rw 0 0
/dev/sda7 /opt/root ext2 rw,usrquota,grpquota 0 0
/dev/sda6 /var ext2 rw,usrquota,grpquota 0 0
/dev/sda1 /boot ext2 rw 0 0
none /dev/pts devpts rw,gid=5,mode=620 0 0
/dev/sda5 /mnt ext2 rw 0 0
a strace on the mount command shows that the syscall mount is executed w/o
error.
mount("/dev/sda5", "/mnt/", "ext2", 0xc0ed0000, 0) = 0
As far as i know it should return EBUSY on a fs that is already mounted.
Second mystery:
the mounted /dev/hda3 contains the same data as /dev/sda5 and changes to one
of them affect both partitions.
I'm not sure wether this is a kernel issue or a escalade firmware issue but
perhaps one of you had that problem too and knows a solution.
Thank you
andreas
--
e-admin internet gmbh
Andreas Gietl
Roter-Brach-Weg 124a
tel +49 941 3810884
fax +49 941 3810891
mobil +49 171 6070008
On Sun, 14 Oct 2001, Andreas Gietl wrote:
> a strace on the mount command shows that the syscall mount is executed w/o
> error.
>
> mount("/dev/sda5", "/mnt/", "ext2", 0xc0ed0000, 0) = 0
>
> As far as i know it should return EBUSY on a fs that is already mounted.
No, it shouldn't. HOWEVER, trees should be in sync (actually, there's
only one tree and only one superblock).
> Second mystery:
>
> the mounted /dev/hda3 contains the same data as /dev/sda5 and changes to one
> of them affect both partitions.
Please, post your /proc/mounts. Results of ls -l /dev/hda3 /dev/sda5
would also be very interesting.