Raid issue

Discussion in 'Server Operation' started by Showa, Nov 15, 2014.

  1. Showa

    Showa Member

    I have been using my Centos server for a couple of years now without any serious problem.
    Three days I had to restart the server, booting stopped at grub. Searching the internet I did try to resolve this with linux rescue. With no result.
    Momentary the os is 5.11 I have got a software raid system with the os in raid 1 on two disks and my data in raid 5 on four disks.
    Using parted magic the disks seems to be ok. The strange thing I noticed was dat md0 (with grub) exists on sde and sdf, md1 md2 and md3 (the other devices in raid 1) are empty instead there are dm-1 dm-2 dm-3. The data md126 (raid 5) became dm-0. Looking at the dm-devices all the files seems to be there.

    Now my my question: is it possible to repair the raid devices with linux rescue and how? Three days of searching didn't resolve my problem and I need the server for my work.

    Thank you in advance,

    Ad.
     
  2. Showa

    Showa Member

    Seems to be a grub problem. I manually added a grub.conf file everywhere I could find grub files using root (hd0,0) I now got the grub menu but hd0,0 is not correct. How can I discover what this has to be?
     
  3. Showa

    Showa Member

    Took root (hd0,0) out of the grub.conf file and everything is working again.
     
  4. mmidgett

    mmidgett Member

    What you have experienced was the draw back to software raid. You have to install grub on every bootable disk to make sure it boots when a disk fails. :)
     

Share This Page