DSM-G600, DNS-3xx and NSA-220 Hack Forum

Mijzelf wrote:

Windows cannot do it, Mac I don't know, but a Linux box can mount that disks, assuming the filesystem and/or partition table is not damaged.

Connect both disks to your computer (directly via sata or with a sata-to-usb converter, it doesn't matter), than open a terminal (command prompt), and type:

Code:

cat /proc/partitions

You should be able to recognize your both raid0 parts. They are probably sdb2 and sdc2. Assuming they are sdb2 and sdc2, and that you are running Ubuntu:
Install mdadm:

Code:

sudo apt-get update
sudo apt-get install mdadm

Assemble the raid array:

Code:

sudo mdadm --assemble /dev/md0 /dev/sdb2 /dev/sdc2

Create a mountpoint and mount:

Code:

sudo mkdir /mnt/raid0
sudo mount /dev/md0 /mnt/raid0

When everything succeeds, you can find your data in /mnt/raid0

Thanks a million, this worked like a charm. I had few errors but after fixing them I was able to mount it and almost (99%) all the files are there intact.
Wojtek

I will list my setup for you. I hope this may inspire you to not use raid. I have a 323 with 2 2TB drives in it. I have each drive configured as an individual drive. I therefor have 4TB of room to play with. This is incredibly excessive since I only have 100GB of monthly bandwidth so it would take 40months to fill it up. Regardless.. I did not set it up this way to get 4TB of storage. I did it for backup purposes. Raid 1 is meant to prevent loss of uptime for critical storage. If you lose a HD, you can ideally replace it, rebuild the array and never have any downtown. For home use this is silly. It provides no backup. If you accidently delete a file.. it's gone. Raid is not a backup solution. So what I use instead is to have all the data from one drive sync'd with the data on the other drive. This is good for a few reasons. I have it sync once per day(using rsync) so if i delete a file accidently, I have until the next sync, to get the file from the backup folder. If I lose a HD, it is as easy as removing the bad hd. I can replace the HD and resync it without issue. If you are really paranoid about your working folder on the nas, you could even schedule rotating 7-day backups so you'll always have a backed up copy up for any day in the past 7.

anyway.. I know almost everyone on here uses RAID1 in their nas.. I just wanted to let you know that there are better solutions.

VirTERM wrote:

This is another way. Thanks. After I run fsck on my drives they actually started working within DNS-323, so I did not loose anything, well I've lost a day .
My DNS-323 acts as a media storage for 2 of my Popcorn Hour media players. In addition, I have 1TB hard drive in each player. Since I only have movies on the DNS-323 it's not critical and I decided to keep RAID-0 as some potential performance boost is always welcome. I just re-arranged things around and new stuff (not seen) does not go there.

There is not performance boost at all. The bottleneck of the DNS-323 is not the HDD. So i'd recommend RAID1 or individual drives, too.

Last edited by oxygen (2010-08-23 18:36:44)

Mijzelf wrote:

Windows cannot do it, Mac I don't know, but a Linux box can mount that disks, assuming the filesystem and/or partition table is not damaged.

Connect both disks to your computer (directly via sata or with a sata-to-usb converter, it doesn't matter), than open a terminal (command prompt), and type:

Code:

cat /proc/partitions

You should be able to recognize your both raid0 parts. They are probably sdb2 and sdc2. Assuming they are sdb2 and sdc2, and that you are running Ubuntu:
Install mdadm:

Code:

sudo apt-get update
sudo apt-get install mdadm

Assemble the raid array:

Code:

sudo mdadm --assemble /dev/md0 /dev/sdb2 /dev/sdc2

Create a mountpoint and mount:

Code:

sudo mkdir /mnt/raid0
sudo mount /dev/md0 /mnt/raid0

When everything succeeds, you can find your data in /mnt/raid0

Hi,

It doesn't worked for me.
I wasn't able to access date correctly on the NAS (maybe bad sectors or other)
I have tried to make a firmware update, from 1.03 to 1.09.
But I didn't know that the Raid 0 configuration will be lost.

So I tried your method in order to restore the raid configuration, but after the command "sudo mdadm --assemble /dev/md0 /dev/sdb2 /dev/sdc2", I've got the message "/dev/sdc2 has no superblock".
Can I re-create this superblock in order to re-assemble the raid array?

Can you help me about that?
Thanks a lot.

(Sorry for my English, I'm not fluent)

Mijzelf wrote:

@leblueboy: It seems there is some damaged data on sdc. *Maybe* you can build the array instead of assembling it.

Code:

sudo mdadm --build /dev/md0 --level=raid0 --raid-devices=2 /dev/sdb2 /dev/sdc2

Fot safety, I'd mount this array readonly (at least the first time)

Code:

sudo mkdir /mnt/raid0
sudo mount -o ro /dev/md0 /mnt/raid0

I've already done a build, but when I mounted the raid, I didn't use the "-o ro" option.
And when I do the "sudo du -shc /mnt/raid0" command, the response is :
20k /mnt/raid0
20 total

(I tried with your command)

Have you another solution to solve my problem?
Thanks

I don't know whether I lost my data due to a power surge or updating from version 1.04 to 1.09 as recommended by DLINK support.  They assured me that I wouldn't lose my data.  My system used to show as RAID but is now showing as Standard and both volumes are 0.

From reading the previous posting on this issue, it sounds like it can possibly be restored, but I'm not very knowledgable in this area and I really need to be able to restore the data lost (personal/business).  There are alot of terminologies used in this forum that I don't understand.  Would you recommend I take the time to learn the terms or simply hire someone to come in and try to restore my lost data?

Any help is appreciated!