software RAID repair (naprawa macierzy RAID)

1. In this example RAID  contents two disks (/dev/sda, /dev/sdb)  and 4 partitions ( md0, md1, md2, md3)
cat /proc/mdstat

Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4] [multipath]
md1 : active raid1 sdb5[1] sda5[0]
 20979264 blocks [2/2] [UU]

md2 : active raid1 sda6[0] sdb6[1]
 20979968 blocks [2/2] [UU]

md3 : active raid1 sdb7[1] sda7[0]
 880264320 blocks [2/2] [UU]

md0 : active raid1 sdb1[1] sda1[0]
 20979648 blocks [2/2] [UU]

unused devices: <none>

2. If the RAID fail you will see

cat /proc/mdstat
 Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4] [multipath]
 md1 : active raid1 sda5[0]
       20979264 blocks [2/1] [U_]
 md2 : active raid1 sda6[0]
       20979968 blocks [2/1] [U_]
 md3 : active raid1 sda7[0]
       880264320 blocks [2/1] [U_]
 md0 : active raid1 sda1[0]
       20979648 blocks [2/1] [U_]
 unused devices:

3. Disconnect failed hdd

4. Connect new hdd

5. Loook into dmesg for hdd letter

6. remove failed hdd from raid (usuniecie uszkodzonego dysku z macierzy)

mdadm /dev/md0 --fail detached --remove detached
mdadm /dev/md1 --fail detached --remove detached
mdadm /dev/md2 --fail detached --remove detached
mdadm /dev/md3 --fail detached --remove detached

7. copy partitions structure from old GOOD HDD to the new one ( kopiujemy strukture partycji z jednego dysku na drugi)

sfdisk -d /dev/sda | sfdisk /dev/sde

8. Add new hdd into RAID (# dodajemy nowo wsadzony dysk)

mdadm --add /dev/md0 /dev/sde1
mdadm --add /dev/md1 /dev/sde5
mdadm --add /dev/md2 /dev/sde6
mdadm --add /dev/md3 /dev/sde7

9. Rebuilding starts automaticaly, You can see RAID status by

cat /proc/mdstat

 10. and WHEN ALL PARTITIONS ARE REBUILDED run:

mdadm --detail --scan >> /etc/mdadm.conf

raid1 on existing filesystem -> Example

Existing filesystem is on /dev/sda

/dev/root  /
/dev/sda5 /var
/dev/sda6 /usr
/dev/sda8  /home

Connect new disk (for example /dev/sdb)

Copy partition structure from sda to sdc:

sfdisk -d /dev/sda | sfdisk /dev/sdb

or if disk has differ size

create the same size ( or bigger) partitions on sdb

end change type for fd (Linux raid autodetect)

change disk’s blkid

tune2fs -U random /dev/sdb

Next, create the single-disk RAID-1 array. Note the “missing” keyword is specified as one of our devices. We are going to fill this missing device later.

mdadm --create /dev/md0 --level=1 --raid-devices=2 missing /dev/sdb1

or

mdadm --create /dev/md0 --metadata=0.90 --level=1 --raid-devices=2 missing /dev/sdb1

and for other partitions

mdadm --create /dev/md1 --metadata=0.90 --level=1 --raid-devices=2 missing /dev/sdb5

mdadm --create /dev/md2 --metadata=0.90 --level=1 --raid-devices=2 missing /dev/sdb6

mdadm --create /dev/md3 --metadata=0.90 --level=1 --raid-devices=2 missing /dev/sdb7

cat /proc/mdstat shows just created raid

Use the file system of your preference here. I'll use ext4 for this guide.

mkfs -t ext4 -j -L RAID-ONE /dev/md0

mkfs -t ext4 -j -L RAID-ONE /dev/md1

mkfs -t ext4 -j -L RAID-ONE /dev/md2

mkfs -t ext4 -j -L RAID-ONE /dev/md3

Make a file system on the swap partition:

mkswap -L NEW-SWAP /dev/sdb2

The new RAID-1 array is ready to start accepting data! So now we need to mount the array, and copy everything from the old system to the new system

 mkdir /tmp/md0 /tmp/md1 /tmp/md2 /tmp/md3

 mount /dev/md0 /tmp/md0

 mount /dev/md1 /tmp/md1

 mount /dev/md2 /tmp/md2

 mount /dev/md3 /tmp/md3

 rsync -avxHAXS --delete --progress /home/ /tmp/md3

 rsync -avxHAXS --delete --progress /var/ /tmp/md2

 rsync -avxHAXS --delete --progress /usr/ /tmp/md1

 rsync -avxHAXS --delete --progress / /tmp/md0

edit /etc/fstab and change mounting points to appropriate /dev/md0 .. /dev/md3

and add line:

/dev/sdb2  swap swap defaults 0 0

reboot with opition  root=/dev/md0 (where my root filesystem was located).

telinit 1 ?

First, open /dev/sda with fdisk and change all the partittions you want to have added to the array to type fd – linux raid autodetection.

Then, for each degraded array, add the appropriate non-array device to it:

mdadm /dev/md0 -a /dev/sda1

(wait for finish recovery)

(umount/dev/sda5)

mdadm /dev/md1 -a /dev/sda5

(wait for finish recovery)

umount/dev/sda6

mdadm /dev/md2 -a /dev/sda6

(wait for finish recovery)

umount/dev/sda8

mdadm /dev/md3 -a /dev/sda8

(wait for finish recovery)

Ensure your /etc/lilo.conf has the correct setup:

boot=/dev/md0
raid-extra-boot=mbr-only

root=/dev/md0

and type:

lilo

LILO will write boot information to each of the individual raid devices boot sectors, so if either /boot or your root partition are on failed disks, you’ll still be able to boot.

Create /etc/mdadm/mdadm.conf.
in /etc/mdadm/mdadm.conf put line:
DEVICE /dev/sda* /dev/sdb*

and WHEN ALL PARTITIONS ARE REBUILDED run:

mdadm --detail --scan >> /etc/mdadm.conf