Hi, I've got a buffalo 1TB DriveStation run in RAID 1 mode, i.e. 500G space available, but that beast is preformatted in vfat. Buffalo doesn't recommend to run it as ext2/ext3, so I mount a vfat drive wich has the limitations "no symlinks, no uids, no gids, etc.". As I just want to use the DriveStation as a backup target, I did the following: I created files of 4G size with 'dd if=/dev/zero of=x01-backup.ext3 bs=1G count=4' (as 4GB is the maximum file size under vfat) and formatted them as ext3. I can mount them and work with them, no prob. Now I'd like to have my 25 files of 4GB file size each coherently mounted as one 100G drive. Is this possible? Must be something like Raid 0. TIA Michael
On Sunday 14 October 2007, Michael Kress wrote:
Hi, I've got a buffalo 1TB DriveStation run in RAID 1 mode, i.e. 500G space available, but that beast is preformatted in vfat. Buffalo doesn't recommend to run it as ext2/ext3, so I mount a vfat drive wich has the limitations "no symlinks, no uids, no gids, etc.". As I just want to use the DriveStation as a backup target, I did the following: I created files of 4G size with 'dd if=/dev/zero of=x01-backup.ext3 bs=1G count=4' (as 4GB is the maximum file size under vfat) and formatted them as ext3. I can mount them and work with them, no prob. Now I'd like to have my 25 files of 4GB file size each coherently mounted as one 100G drive. Is this possible? Must be something like Raid 0. TIA Michael
You might try something like this: http://osdir.com/ml/linux.lvm.general/2003-10/msg00096.html
Personally I think it's a bad idea. :)
Shawn
Shawn Everett wrote:
On Sunday 14 October 2007, Michael Kress wrote:
I can mount them and work with them, no prob. Now I'd like to have my 25 files of 4GB file size each coherently mounted as one 100G drive. Is this possible? Must be something like Raid 0. TIA Michael
You might try something like this: http://osdir.com/ml/linux.lvm.general/2003-10/msg00096.html
Personally I think it's a bad idea. :)
Shawn
Hi, I dont't like it either, but I have to adopt to the given facts. :-( Your proposed method seems to be yet elegant. Anyways, I got a problem with it.
This is what I do: -----8<-----8<-----8<-----8<-----8<-----8<-----8<-----8<----- dd if=/dev/zero of=fake-disk-1.img bs=1M count=10 dd if=/dev/zero of=fake-disk-2.img bs=1M count=10 losetup /dev/loop0 fake-disk-1.img losetup /dev/loop1 fake-disk-2.img pvcreate /dev/loop0 pvcreate /dev/loop1 vgcreate /dev/vgtest /dev/loop0 /dev/loop1
root@x05:/root [0] > vgs VG #PV #LV #SN Attr VSize VFree vgtest 2 0 0 wz--n- 16.00M 16.00M
lvcreate --size=16M --name=lvtest /dev/vgtest mkfs /dev/vgtest/lvtest mkdir /mnt/lvtest mount /dev/vgtest/lvtest /mnt/lvtest/
root@x05:/root [0] > df -h Filesystem Size Used Avail Use% Mounted on /dev/sda1 3.0G 2.4G 617M 80% / none 256M 0 256M 0% /dev/shm /dev/mapper/vgtest-lvtest 16M 140K 15M 1% /mnt/lvtest
umount /mnt/lvtest/ shutdown -r now -----8<-----8<-----8<-----8<-----8<-----8<-----8<-----8<-----
So far so good. After reboot I found this:
-----8<-----8<-----8<----- root@x05:/root [0] > lvs No volume groups found root@x05:/root [0] > vgs No volume groups found root@x05:/root [0] > pvs (no output) -----8<-----8<-----8<-----
So I tried to recover the 2 files: I had to 'losetup' the files again, but pvcreate hangs here: -----8<-----8<-----8<-----8<-----8<-----8<-----8<-----8<----- root@x05:/root [0] > pvcreate /dev/loop0 Can't initialize physical volume "/dev/loop0" of volume group "vgtest" without -ff root@x05:/root [5] > -----8<-----8<-----8<-----8<-----8<-----8<-----8<-----8<-----
So I did: -----8<-----8<-----8<-----8<-----8<-----8<-----8<-----8<----- root@x05:/root [5] > pvcreate -ff /dev/loop0 Really INITIALIZE physical volume "/dev/loop0" of volume group "vgtest" [y/n]? y WARNING: Forcing physical volume creation on /dev/loop0 of volume group "vgtest" Physical volume "/dev/loop0" successfully created root@x05:/root [0] > pvcreate -ff /dev/loop1 Couldn't find device with uuid 'Z4Dr2Q-CBeM-QDHP-26d2-O1GV-DgZh-C7Z6pK'. Couldn't find all physical volumes for volume group vgtest. Couldn't find device with uuid 'Z4Dr2Q-CBeM-QDHP-26d2-O1GV-DgZh-C7Z6pK'. Couldn't find all physical volumes for volume group vgtest. get_pv_from_vg_by_id: vg_read failed to read VG vgtest Physical volume "/dev/loop1" successfully created root@x05:/root [0] > vgcreate /dev/vgtest /dev/loop0 /dev/loop1 Volume group "vgtest" successfully created root@x05:/root [0] > lvcreate --size=16M --name=lvtest /dev/vgtest Logical volume "lvtest" created root@x05:/root [0] > mount /dev/vgtest/lvtest /mnt/lvtest/ mount: you must specify the filesystem type root@x05:/root [32] > fsck /dev/vgtest/lvtest fsck 1.35 (28-Feb-2004) e2fsck 1.35 (28-Feb-2004) Couldn't find ext2 superblock, trying backup blocks... Resize inode not valid. Recreate<y>? yes
/dev/vgtest/lvtest was not cleanly unmounted, check forced. Pass 1: Checking inodes, blocks, and sizes Pass 2: Checking directory structure Pass 3: Checking directory connectivity Pass 4: Checking reference counts Pass 5: Checking group summary information Free blocks count wrong for group #0 (7854, counted=7855). Fix<y>? yes
Free blocks count wrong (15722, counted=15723). Fix<y>? yes
/dev/vgtest/lvtest: ***** FILE SYSTEM WAS MODIFIED ***** /dev/vgtest/lvtest: 11/4096 files (0.0% non-contiguous), 661/16384 blocks root@x05:/root [1] > mount /dev/vgtest/lvtest /mnt/lvtest/ root@x05:/root [0] > -----8<-----8<-----8<-----8<-----8<-----8<-----8<-----8<-----
Ok, this works, but I can't rely on such a file system. Any idea on how I could get back the original logical volume lvtest without the indeed ugly file system check and without endangering the data contained in the lv? TIA Michael
-
Michael Kress -
Shawn Everett