CentOS 6.8 x86_64 Software RAID-1 Boot Failure

Discussion in 'Installation/Configuration' started by GenePoole, Feb 14, 2017.

  1. GenePoole

    GenePoole New Member

    Let's start by telling you about the physical environment:
    ASrock AMD motherboard w/ AMD Athlon II 6-core processor
    32 GB (4 x 8GB) RAM
    6 hard drives (4 - 1TB and 2 - 2TB Western Digital)

    Software environment:
    This machine is at least 3+ years old. I started out using CentOS 6.1 and implemented software raid-1 during installation along with LVM. As is required, I set up /boot as a standard EXT3 partition which was /dev/md0 (/dev/sda1 - /dev/sdb1) 1024MB in size. I then configured the required file systems using RAID-1 and lvm as vg_opsys = /dev/md1 (/dev/sda2 - /dev/sdb2) 65536MB and vg_swap = /dev/md2 (/dev/sda3 - /dev/sdb3). I left the remainder of the hard drives to be configured manually after the initial installation.

    Over time many changes were made, hard drives and RAM were added including a Oracle 11g database with 4 instances; KVM with 5 virtual servers.
    On 02/11 I did a normal maintenance process (yum check-update then yum -y update) and since this added a new kernel I did a reboot (shutdown -r now). It would not reboot but only gave me a grub prompt. I rebooted again and it showed me 1 less hard drive than what was installed (normal /dev/sda - /dev/sdf; after reboot /dev/sda - /dev/sde). What was /dev/sdb is now /dev/sda.

    Why did it change the disk designation?
    Is it normal to loose a disk without warning?
    Will I be able to get my data back?


Share This Page