Software Raid0 problem with 5TB

Max Kipness max at kipness.com
Wed Mar 10 18:13:34 UTC 2004


Hello,

I've got a Fedora Core 1 w/2.6.3 kernel with 3 3Ware 8-port cards, with
8 250gb drives each in HW Raid 5 containers. My goal was then to use
software Raid0 to create one volume out of the 3. So far the volume is
not showing the correct total. Here are the steps I've taken.

Oh, my boot drives are also two 80Gb drives using a software Raid1,
which is working fine and I don't thing would have anything to do with
the issues I'm having.

1) Created partitions with fdisk and changed ID to fd on all 3
containers. Doing an fdisk -l gives me the following results which
appear to be correct:

 Disk /dev/sda: 1756.9 GB, 1756994011136 bytes
255 heads, 63 sectors/track, 213609 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot    Start       End    Blocks   Id  System
/dev/sda1             1    213609 1715814261   fd  Linux raid autodetect

Disk /dev/sdb: 1756.9 GB, 1756994011136 bytes
255 heads, 63 sectors/track, 213609 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot    Start       End    Blocks   Id  System
/dev/sdb1             1    213609 1715814261   fd  Linux raid autodetect

Disk /dev/sdc: 1756.9 GB, 1756994011136 bytes
255 heads, 63 sectors/track, 213609 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

2) Configured /etc/raidtab as follows and then issued the command:
mkraid /dev/md1, which ran successfully.

raiddev             /dev/md1
raid-level                  0
nr-raid-disks               3
chunk-size                  32
persistent-superblock       1
nr-spare-disks              0
    device          /dev/sda1
    raid-disk     0
    device          /dev/sdb1
    raid-disk     1
    device          /dev/sdc1
    raid-disk     2

3) Mounted md1 with ext3 successfully

But when doing an df -h, I get the following bizarre results:

Filesystem            Size  Used Avail Use% Mounted on
/dev/md0               75G  1.4G   69G   2% /
none                 1008M     0 1008M   0% /dev/shm
/dev/md1              801G   33M  760G   1% /data

I have no idea where it is getting 801 GB instead of something like 5.1
TB. 

To experiment, I used only two of the 1.7TB partitions, and this seemed
to work fine. Here are the results of df -h:

Filesystem            Size  Used Avail Use% Mounted on
/dev/md0               75G  1.9G   69G   3% /
none                 1013M     0 1013M   0% /dev/shm
/dev/md1              3.2T   33M  3.0T   1% /data

This tells me that large block device support is working. But why won't
it allow the 3rd 1.7Tb partition? Am I hitting a limit with Fedora? With
Raidtools?

Any help would be much appreciated.

Max 





More information about the fedora-list mailing list