Raid Soft : degraded array (resolved)

**didi150** · 07-20-2006

Hello all,

Following a big power failure my computer crashed, (the motherboard was dead).
Once the motherboard changed, I saw that my raid 1 array (soft) was noted as "degraded"

I have resolved this problem, and as it is not really easy to repair (I wasn't able to find info on google), I wanted to post something in order to help others in similar case. (google will index this page I hope)

The problem :

In the logs I saw thinks like

Code:

md: kicking non-fresh hdc1 from array!

kernel: ide1: BM-DMA at 0xffa8-0xffaf, BIOS settings: hdc:DMA, hdd:DMA
kernel: hdc: max request size: 512KiB
kernel: hdc: 398297088 sectors (203928 MB) w/8192KiB Cache, CHS=24792/255/63, UDMA(133)
kernel: hdc: cache flushes supported
kernel: hdc: hdc1
kernel: md: bind<hdc1>
kernel: md: unbind<hdc1>
kernel: md: export_rdev(hdc1)

My array is composed of /dev/hda1 and /dev/hdc1 to form a /dev/md0
This /dev/md0 is used enterely for lvm2 logical volumes. My slash is on the array and also my /boot/

The second part of the array was unable to be resync at boot

Code:

mdadm --query --detail /dev/md0

Showed something like

Code:

/dev/md0:
        Version : 00.90.03
  Creation Time : Sat Jan 14 14:35:17 2006
     Raid Level : raid1
     Array Size : 199141632 (189.92 GiB 203.92 GB)
    Device Size : 199141632 (189.92 GiB 203.92 GB)
   Raid Devices : 2
  Total Devices : 2
Preferred Minor : 0
    Persistence : Superblock is persistent

    Update Time : Thu Jul 20 11:14:57 2006
          State : clean
 Active Devices : 1
Working Devices : 1
 Failed Devices : 0
  Spare Devices : 0

           UUID : 1a94bb53:1c629336:3129889f:fb3064cb
         Events : 0.12489983

    Number   Major   Minor   RaidDevice State
       0       3        1        0      active sync   /dev/hda1
       1       0        0        0      removed

The second disk /dev/hdc was "removed" but not failed

I tried to hot add /dev/hdc1 to the array by typing

Code:

mdadm /dev/md0 --add /dev/hdc1

But the system answered

Code:

Cannot open /dev/hdc1 : Device or ressource busy

So I though lvm was maybe in cause... and it was right

As the /dev/hdc1 was unable to enter the array, the /dev/hdc1 was available to lvm
(In the boot sequence there is "vgchange -a y vg0")

So the device was busy by the kernel (that's why lsof showed nothing in use on /dev/hdc1)

So, I open my initrd image, mount it, copy it to be able to modify it

In the "script" file /mnt/initrd17tip-rw/script
You can see how the kernel is assembling the array at boot time

Code:

mdadm -A /dev/md0 -R -u 1a94bb53:1c629336:3129889f:fb3064cb /dev/hda1 /dev/hdc1

So I add

Code:

mdadm /dev/md0 --add /dev/hdc1

just after the assembling line that was failing

I recreate the initrd image by doing :

Code:

mkcramfs initrd17tip-rw /boot/initrd.img-2.6.17.1tip1

Rerun lilo and reboot

after reboot, the array was reconstructing,

Code:

mdadm --query --detail /dev/md0

Showed something like

Code:

    Number   Major   Minor   RaidDevice State
       0       3        1        0      active sync   /dev/hda1
       1       0        0        0      removed
       2       ?        22      1      spare syncing /dev/hdc1

Once the resync was ok, the mdadm show only two device synced

I really hope it will help someone

**forgottentq** · 07-24-2006

something extremely similar happened to me yesterday.. this really helped! thanks.

Thread Tools

Display

Raid Soft : degraded array (resolved)

Posting Permissions