Partitionable Raid

Home » CentOS » Partitionable Raid
CentOS 6 Comments


After reading the tutorial at I have the following question:

What should I put instead of splashimage=(hd0,0)/grub/splash.xpm.gz


root (hd0,0)

on /etc/grub.conf? Should I leave those lines untouched? If so, how would GRUB know where to boot from if /dev/sda fails?

Or I would need to swap the drives in order to boot properly and then put a new one as /dev/sdb and then add it to the array?

Thanks in advance for all the help.

Best regards,

Eugenio Pacheco

6 thoughts on - Partitionable Raid

  • If sda fails outright, sdb will BECOME sda. This is usually the case when a “drive fails”.

    If sda is skillfully overwritten in specific places with the intention of producing a curcumstance in which Linux kernel raid will fail, then Linux kernel raid will fail.

  • Please steer away from the above if possible. I did this around June/2013 time frame and posted my woes in this mailing list.

    Please search the archives and go through the entire thread.

  • In the case of the partionable RAID stated by OP, the boot fails (see my posting from June/2012.

    Even a non raid will fail, will it not?

  • W dniu 2013-10-02 13:03, Arun Khan pisze:
    Yes, June. But, if i good remember, this was a problem with dracut itself. I have had C6.2 with standard raid-1 at top of partition and suffered this problem too. Dracut update solved my problem with system non booting with one drive in RAID-1 scenario. dracut updated Oct 03 2012.

    So above solution at wiki isn’t bad at all.


  • Indeed, it was a dracut problem but for me, with dracut updates, there was no solution in the Aug/Sep time frame during my consultation gig at client site.

    YMMV and it is good that the problem has been fixed.