Home / server / Questions / 178157
In Process
Ian Burris
Asked: 2010-09-05 14:02:20 +0800 CST

Reinstalling GRUB on RAID array

  • 772

I have a computer running Ubuntu 10.04 Desktop x64. I just bought another tb drive and wanted to combine my two drives using RAID 0. I have a MSI 890GXM-G65 motherboard and decided to use their built in RAID support. The last step of setting up RAID asked for confirmation to delete the MBR and I had a feeling things were going to go wrong but decided to chance it. Needless to say I can no longer boot into Ubuntu. I did some searching and most suggested booting into a live version of Ubuntu and reinstalling GRUB from there. I tried this but can't seem to mount the harddrives. Any ideas on how I'd go about mounting and reinstalling GRUB? Thanks in advance for any help you can offer.

mount raid grub mbr ubuntu-10.04

3 Answers

  1. From http://web.mit.edu/rhel-doc/5/RHEL-5-manual/Deployment_Guide-en-US/s1-raid-config.html: If you are making a RAID partition of /boot/, you must choose RAID level 1, and it must use one of the first two drives (IDE first, SCSI second). If you are not creating a seperate RAID partition of /boot/, and you are making a RAID partition for the root file system (/), it must be RAID level 1 and must use one of the first two drives (IDE first, SCSI second).

    As for mounting the RAID, is the onboard controller compatible with the kernel? Do the devices show under /dev/md*? Is there any mention of RAID in lsmod or dmesg?

    • 1

  2. When you set up a bunch of disks as RAID0, this changes the way the data on the disk is laid out. (Echoing ErikA, if you don't understand this, you have no business using RAID0). Unless your motherboard is more sophisticated than I'd expect from a consumer board (or even an expensive RAID controller), it didn't keep any of your data. (In principle, it could have copied the data from one of the disks, but that's long and a little complicated.)

    After the motherboard created the RAID0, I would expect that from Linux:

    • either you see one big disk with no partitions, if Linux supports RAID0 on this motherboard);

    • or you see no disk or two disks containing garbage, if Linux doesn't support RAID0 on this motherboard.

    If Linux does support your motherboard, you now need to either reinstall the system on the RAID0 array and restore your data from backups, or create a reasonable partition table and restore both system and data from backups. But I don't recommend using hardware RAID with consumer hardware (and rarely with server hardware, but it can be faster for modes that require a checksum calculation): Linux's has reliable and fast RAID0 and RAID1 support.

    If you choose to use Linux's built-in RAID0, turn off RAID0 from the motherboard, then create a RAID0 array spanning both drives with mdadm, and reinstall or restore your backups.

    If you want to back out and not use RAID0 after all, there's a chance that the conversion to RAID0 didn't destroy much of what was on the disk. First turn off RAID0 in the BIOS, then try recreating the partition table.

    • 0

  3. To be blunt, it's apparent you don't know what RAID0 is or what the implications are. Before you made them into a RAID0 pair you had 2 separate disks, now you have one virtual disk on which the data is interleaved between the physical disks. There is no way for the OS to book from the existing install because it's effectively not there any more.

    If your onboard controller supports it, you could try a fresh install and hope that it works, or you can undo the RAID0 configuration and use the buit in mdadm as Gilles suggests to create a software RAID0 (or similar) for the partitions you want.

    • 0