Raid0 assemble disk again?

Question

Our server recently suffered a disk failure, so our hosting provider put a new Disk in with CentOS so we can login and see if we can recover the data.

We had 2 x 120 GB SSD drives in software RAID0 config - our host set this up for us; we didn't do it ourselves as we lack the "know-how" - hence why I am here posting this.

Here's the output of fdisk -l -

SDB and SDC are the RAID drives.

I tried the following commands to mount the drives:

I did some digging around and found that for software RAID, you can use the mdadm tool to scan/assemble automatically, but this didn't work either:

[root@localhost ~]# mdadm -A --scan
mdadm: No arrays found in config file or automatically

I tried running the examine option and this is the output:

[root@localhost ~]# mdadm --examine /dev/sdb
/dev/sdb:
   MBR Magic : aa55
Partition[0] :       204800 sectors at         2048 (type 83)
Partition[1] :     20480000 sectors at       206848 (type 83)
Partition[2] :      8192000 sectors at     20686848 (type 82)
Partition[3] :    435964672 sectors at     28878848 (type 05)
[root@localhost ~]# mdadm --examine /dev/sdb1
mdadm: No md superblock detected on /dev/sdb1.
[root@localhost ~]# mdadm --examine /dev/sdb2
mdadm: No md superblock detected on /dev/sdb2.
[root@localhost ~]# mdadm --examine /dev/sdb3
mdadm: No md superblock detected on /dev/sdb3.
[root@localhost ~]# mdadm --examine /dev/sdb4
mdadm: No md superblock detected on /dev/sdb4.

[root@localhost ~]# mdadm --examine /dev/sdc
mdadm: No md superblock detected on /dev/sdc.
[root@localhost ~]# mdadm --examine /dev/sdc1
mdadm: cannot open /dev/sdc1: No such file or directory

and here's the output of cat /proc/mdstat

root@localhost ~]# cat /proc/mdstat
Personalities :
unused devices: <none>

I had a similar issue before and I already asked about it before - re-mount two old disk from raid0 setup to recover data

Last time I somehow managed to fix it because the array was 100% clean and I managed to mount it by running this command: mkdir /mnt/oldData && mount /dev/md127 /mnt/oldData but this time the issue appears to be different. There is no /dev/md* - see this ->

Can someone help?

Answer 1

As I understand your issue, there was a disk failure in your RAID 0 array. RAID 0 has NO FAULT tolerance and is primarily used for speed. Without both disks, or using an alternate RAID array setup that has parity, you cannot recover this data.

If you're able to recover the data from the disk that failed, you can attempt to move that data to a new drive and add it back to the array. This is not guaranteed to work, but it can be attempted if you still have the failed drive, or if that drive's data is somehow retrieved.

Answer 2

mdadm --examine --metadata 1.2 /dev/sd[bc]

Looks like full block device software raid. Not partition based one - partitions are defined on the first drive.

Try to find the metadata on the disks using various metadata versions. There is possibility, that booted live distro, is not supporting your metadata type.

man mdadm is your friend :)

---

Here is an example with mine system: 2x30GB SSD:

mdadm --examine --metadata 1.2 /dev/sdb1

/dev/sdb1:
Magic : a92b4efc
Version : 1.2
Feature Map : 0x0
Array UUID : c609e03c:bda63289:1c4f3bbe:d35ba549
Name : sh:0 (local to host sh)
Creation Time : Thu Aug 29 16:11:22 2013
Raid Level : raid0
Raid Devices : 2

Avail Dev Size : 62529520 (29.82 GiB 32.02 GB)
Data Offset : 16 sectors
Super Offset : 8 sectors
State : clean
Device UUID : 7c962127:e548bb1c:0c3e4a7a:e180fa57

Update Time : Thu Aug 29 16:11:22 2013
Checksum : 657466b3 - correct
Events : 0

Chunk Size : 512K

Device Role : Active device 1
Array State : AA ('A' == active, '.' == missing)

mdadm --examine --metadata 0.9 /dev/sdb1

mdadm: No super block found on /dev/sdb1 (Expected magic a92b4efc, got 82901084)