6

I had setup two 500gb disk in RAID0 on my server, but recently suffered a hard disk failure (saw a S.M.A.R.T error on the HDD at boot). My host has put 2 new disk in RAID-0 again (re-installed the OS) and re-attached the old drives on the same machine, so that I can recover the data.

My old drives are:

  • /dev/sdb
  • /dev/sdc

How can I mount these two disks back in RAID0, so that we can recover the data from our old drive? Or is this not possible any more? Have I lost all my data?

This is my /etc/fstab and df -h

This is my fdisk -l: 

[root@localhost ~]# fdisk -l

Disk /dev/sda: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00040cf1

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *           1          13      102400   83  Linux
Partition 1 does not end on cylinder boundary.
/dev/sda2              13        1288    10240000   83  Linux
/dev/sda3            1288        2333     8388608   82  Linux swap / Solaris

Disk /dev/sdc: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0005159c

   Device Boot      Start         End      Blocks   Id  System
/dev/sdc1               1       60802   488385536   fd  Linux raid autodetect

Disk /dev/sdb: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0006dd55

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1   *           1          26      204800   83  Linux
Partition 1 does not end on cylinder boundary.
/dev/sdb2              26        4106    32768000   82  Linux swap / Solaris
/dev/sdb3            4106        5380    10240000   83  Linux
/dev/sdb4            5380       60802   445172736    5  Extended
/dev/sdb5            5380       60802   445171712   fd  Linux raid autodetect

Disk /dev/sdd: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x9f639f63

   Device Boot      Start         End      Blocks   Id  System
/dev/sdd1               1       60802   488385536   83  Linux

Disk /dev/md127: 956.0 GB, 955960524800 bytes
2 heads, 4 sectors/track, 233388800 cylinders
Units = cylinders of 8 * 512 = 4096 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 524288 bytes / 1048576 bytes
Disk identifier: 0x00000000

I read somewhere that you can do this with this command: mdadm -A --scan however, it does not yield any result for me -> No arrays found in config file or automatically

Improve this question
11
  • What is the output of mdadm --examine /dev/sdb and mdadm --examine /dev/sdc ?
    – Darth Android
    Apr 29, 2013 at 14:14
  • Also, what is the output of cat /proc/mdstat ?
    – Darth Android
    Apr 29, 2013 at 14:24
  • Just realized, you probably want to mdadm --examine the /dev/sdb5 and /dev/sdc1 partitions instead of the drives themselves.
    – Darth Android
    Apr 29, 2013 at 14:36
  • I've examined it and here's the result: pastebin.com/raw.php?i=nnFzw0GG
    – Latheesan
    Apr 29, 2013 at 21:48
  • Uh, what's the output of cat /proc/mdstat ? That pastebin says that your old array is functioning just fine and is 100% clean - in fact, the reason you can't assemble the array is because it's already assembled. Try mkdir /mnt/oldData && mount /dev/md127 /mnt/oldData. That said, if one of the drives was giving a SMART error, I wouldn't trust the drive any more, and would still back up all of the data off of it.
    – Darth Android
    Apr 29, 2013 at 22:42

1 Answer 1

Reset to default
4

First, determine the value of the data. If this is business-critical data that you have to have, evaluate your options with respect to sending the disks to a professional data recovery service. Self-recovery from dying disks and crashed RAID arrays is always a bit off the edge of the map. If you're already assuming that the data on the old drives is lost and you're just hoping for some recovery of data, and don't want to spend additional money, then proceed.

You will probably have to force the array together. This can result in silent corruption because the RAID knows that it's not clean, and you're telling it to put on a smile and pretend it is anyways. Just keep in mind you will need to manually verify integrity of any files you pull off of the RAID.

You can force an array together with:

mdadm --assemble --force /dev/md126 /dev/sdb5 /dev/sdc1

If /dev/md126 already exists on your system, pick the next one down (/dev/md125) until you find a free (non-existent) device.

This should force the array into a working state. Let's mount the filesystem read-only so that we can pull data off of it without corrupting anything further

mkdir /mnt/oldData
mount /dev/md126 /mnt/oldData -o ro

At this point you should be able to copy data out of /mnt/oldData and to a safe location.

Improve this answer
1
  • 1
    This is exactly what I want to do right now; recover the data. I tried your reccomendation and I get the following error: mdadm: /dev/sdb5 is busy - skipping and mdadm: /dev/sdc1 is busy - skipping. I used mdadm to --examine sdb5 and sdc1 and this is what it says: pastebin.com/raw.php?i=nnFzw0GG
    – Latheesan
    Apr 29, 2013 at 21:47

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .