Mount NTFS ON RAID 0 (Fedora 10)

Sabba Hillel · 02-24-2009, 02:57 PM

I have had a similar problem with attempting to build Fedora 10 on a Dell PowerEdge with a hardware RAID configuration. I installed Fedora 10 using a complete install (not an upgrade) and initialized the partitions completely. It appears to build correctly from the Fedora 10 DVD. However, after the end of the installation, the system reboot will give the error message

Volume Group "VolGroup00" not found
Unable to access remote device (/dev/VolGroup00/LogVol01)
mount error mounting /dev/root on /sysroot as ext3: no such file or directory

At this point, it locks up and I cannot even get into grub.

The problem seems similar and may point to the difficulty here as well. Fedora 7 and Fedora9 both install and boot properly. Does anyone know why Fedora 10 would no longer build?

mostlyharmless · 03-02-2009, 02:49 PM

!!!!!, In case you are still interested:

Your sda and sdb are raid0 together. The second disk in a firmware raid0 set gets reported as having no partition table by fdisk, which is what you see. The first disk has the table for both, but the numbers may appear screwy. Don't worry about it, just don't mess with the partition table at this point.

Your raid is firmware raid or "fakeraid". It has already been setup by dmraid as listed by you:

Quote:

control via_cgddjifhah via_cgddjifhahp1 via_cgddjifhahp2 via_cgddjifhahp3 via_cgddjifhahp4 via_cgddjifhahp5 via_cgddjifhahp6

which shows that you have a via Raid controller. Typing dmraid -s will show you the raid set.

As Quakeboy02 pointed out, you need merely mount the appropriate partitions such as /dev/mapper/via_cgddjifhahp1 using mount and the ntfs or ntfs-3g filesystem. eg.

Quote:

mount -t ntfs /dev/mapper/via_cgddjifhahp1 /mnt

Sabba Hillel: you have a different problem and probably need your own thread.

!!!!! · 07-18-2009, 02:39 PM

mostlyharmless....you did it man,actually I did this before (not me,but a friend works in our data centre but he didn't told me how he did it.)now i did the last command you said and i can see my windows files.thanks a lot.

!!!!! · 07-19-2009, 01:37 PM

unfortunately the previous mount ntfs files temporary but everything disappear after reboot.

mostlyharmless · 07-20-2009, 02:44 PM

Put the command in your rc.local file (usually in /etc/rc.d) and it will be executed every time the system boots.

!!!!! · 07-25-2009, 07:10 AM

i will try do this,but someone told me that i must edit something in the fstab do you have any idea how that done.?

mostlyharmless · 07-27-2009, 10:31 AM

You could probably put it in fstab instead, but it is not necessary.

There's nothing wrong with just putting the

mount -t ntfs /dev/mapper/via_cgddjifhahp1 /mnt

(or preferably)

mount -t ntfs-3g /dev/mapper/via_cgddjifhahp1 /mnt

at the end of your rc.local using pico. You have to be root to mount or edit rc.local.

PTrenholme · 07-27-2009, 03:31 PM

After you've done a sudo mkdir /Vista and sudo mkdir /Vista/Recovery you could do a sudo nano /etc/fstab and add this:

Code:

UUID=342D7D7923AC4107   /Vista                  ntfs-3g rw,defaults     0 0
UUID=2C88743C8874071C   /Vista/Recovery         ntfs-3g rw,defaults     0 0

where the numbers after the UUID= can be found in the output of ls -l /dev/disk/by-uuid/

Note that the mount point names are almost completely arbitrary, and can by in any directory you want to place them. I like to mount hard drives directly under / with name that start with an upper-case letter, but that's not a standard *nix convention. As mostelyharmless has suggested, the "conventional" mount point is a directory created under /mnt. (Although one way to read his suggestion is to mount the drive on /mnt which would preclude mounting more than one drive.) Bottom line: Where you decide to mount your NTFS drives is entirely up to you.