Recovering data from Lacie Big Disk
March 29th, 2009
I’ve had a Lacie Big Disk Ethernet 1TB for some time now. I used it as a mediaserver but since it’s not built to handle so many files (20.000 + mp3) the boot disk got filled up and didnt have any space left leaving it unable to boot, leaving me fucked and unable to access 700GB of data and backups.
Since Lacie didn’t have a solution I decided to find my own to recover my data. And here it is.
How to recover your data
Few facts about the Lacie Big Disk Ethernet (1TB)
- It is Using a custom built Linux to run
- It uses JBOD (linear / span) to build a 1TB disk out of two 500GB.
You need Ubuntu for this since the Lacie Big Disk Ethernet runs on Linux. I’m running mine on VMware Fusion on Mac.
1. Get a dual SATA casing (I’m using NexStax MX)
2. Put your disk inside the casing. DO NOT SET IT TO JBOD MODE! Just standalone disks.
3. When you turn it on DO NOT INITIALIZE THE DISKS. You will damage your data.
4. Open Ubuntu and connect the disks.
5. Open up Terminal and Install mdadm => sudo apt-get -install mdadm
6. Connect your device.
7. Three drives should mount up
8. Open terminal and type => sudo fdisk -l <= You should get alist of devices.
9. Find the correct paths to your disks.
- The first disk in my case was=> /dev/sdb2 487379970 (blocks)
- The second one was a 500GB disk => /dev/sdc <= Note: This one returns a “doesn’t contain a valid partition table” error.
10. Open up /etc/mdadm.conf under the 130mb mount.
11. Open up Terminal and type => sudo nano /etc/mdadm/mdadm.conf
12. Copy the settings from the Lacie config in step 10 to the config file in step 11. Change the paths to match your current setup. The smaller mount goes first and the bigger one (without the partition ) second. Like so:
- DEVICE /dev/sdb2 /dev/sdc
- ARRAY /dev/md0 level=linear num-devices=2 UUID=163218f3:299b9ccc:3c666c38:6053
13. Hit CTRL-X and press Y to save the changes.
14. Now in terminal type => sudo mdadm—assemble /dev/md0
15. Whualla! Your data, safe and ready to be copied to a nother drive!
Note that if your config has a different /dev/mdx path then mount that one up.
Hope this helps someone to get their data back!
Henna
March 31st, 2009 at 03:08 PMLink to this comment
Thanks Snorri! This is exactly what I needed, to see that AM NOT A TOTAL NERD. You are tho! hugglez
Snorri Páll
March 31st, 2009 at 04:55 PMLink to this comment
haha evil evil bastard
Eric
April 15th, 2009 at 06:49 AMLink to this comment
Thank for your article. My Lacie 1TB Big Disk is dead. After connecting the 2 SATA disks to my Ubuntu 8.10, and copied the mdadm.conf to my ubuntu, I did a “sudo mdadm—assemble /dev/md0 /dev/sda2 /dev/sdb”. It came back saying “mdadm: no RAID superblock on /dev/sdb, mdadm: /dev/sdb has no superblock – assembly aborted”. Then I did a “sudo mdadm—examine /dev/sda2”, it showed me all the RAID information. However, “sudo mdadm—examine /dev/sdb”, it came back “mdadm: No md superblock detected on /dev/sdb”. Do you have the same result, or my /dev/sdb is dead? Thanks.
Eric
April 16th, 2009 at 05:13 AMLink to this comment
I used another command: “sudo mdadm—build /dev/md0—level=linear—raid-devices=2 /dev/sda2 /dev/sdb”, and it worked. Hooray.
Damien
April 20th, 2009 at 06:46 AMLink to this comment
I am looking to do this as well. Can I just use my two onboard SATA Connectors and boot Linux from IDE?
Eric
April 27th, 2009 at 11:24 PMLink to this comment
Yes, I think so. Just make sure the two disks from Lacie are /dev/sda (with partitions) and /dev/sdb in Linux.
Snorri Páll
April 28th, 2009 at 08:29 PMLink to this comment
Sorry for the late reply.
Eric: See you figured this out. Glad the article helped someone.
Damien: Yes that should work. Good luck :)
Ken
June 3rd, 2009 at 04:56 AMLink to this comment
Hi, Thanks for this resource! My Ethernet Big Disk died and I got an RMA, but they won’t return the data so I thought I was screwed.
I’m a total linux noob, but I think I have the array back. /dev/md0 is built and ready.
Okay now what? Do I mount it? How?
snorri
June 7th, 2009 at 09:14 PMLink to this comment
not the linux nerd myself, just managed to put the bits together so save my data.
did you try using mdadm-assemble? that should mount the volume if the configuration is correct.
sudo mdadm—assemble /dev/md0
snorri
June 7th, 2009 at 09:15 PMLink to this comment
Also if its a RMA.. are you sure your data is still there?
Ken
June 9th, 2009 at 05:08 PMLink to this comment
Thanks for the help. I really appreciate it. I tried both assemble and build, and build seemed to work. I can see the backend linux shares from the disk, just not the data. No, I’m not sure the data is still there. The unit just wouldn’t turn back on one day. I had it happen once before and it was a faulty power supply. They sent me a new one, but this time it wasn’t the power supply.
I can see an 8.2MB share, a 131.6MB share and a 748.5MB share, but that’s all. If I can’t see the data, then presumably it’s gone?
Tim
June 13th, 2009 at 09:45 AMLink to this comment
Thanks for the Data recovery howto. I have a different problem, I’ve bought a device with no disks in it. I dont suppose you’d be able to dd images of the OS partitions for me, like has been done on http://gpl.nas-central.org/LACIE/ethernet_disk_mini_v2_-_ethernet_big_disk/EDmini_v2_partition_images/.
I’ve managed to get the device 1/2 working with the EDmini OS. Cheers
Micheil Keegan
June 17th, 2009 at 08:53 AMLink to this comment
I’m having a lot of trouble getting this to work. I’m using a live distro of SUSE as I don’t have linux installed. I have the two drives on /dev/sda and /dev/sdc.
mdadm says /dev/md0 has been started with 2 drives. where do i find the mount? or the 2 drives?
Any help would be appreciated.
Snorri
June 19th, 2009 at 10:56 AMLink to this comment
Ken: As i said im no linux genie, just managed to figure this out and the data mount just automounted in Ubuntu (8.x) when i ran sudo mdadm—assemble /dev/md0
Also the linux distro you use needs to support XFS since the datamount is XFS type of filesystem.
Tim: I recommend you get a different casing, like Drobo or somthing since Lacie Big Disk is shit => http://snorripall.com/not-gonna-use-it-dont-fucking-make-it
Micheil Keegan: Are you sure SUSE supports XFS file system? the datamount is in XFS format, might be the reason why it isnt mounting up correctly. Ubuntu 8.x supports XFS.
Harry
September 2nd, 2009 at 06:56 PMLink to this comment
I wish I had a clue how to carry all that out. I don’t even know what a console is. I’m not a complete idiot but have no experience with linux at all, other than my failed lacie ethernet big disc.
Harry
September 3rd, 2009 at 04:36 AMLink to this comment
Alright I may need some help with this. I am typing from Ubunto right now. Like I said, I’m not an idiot. In fact I’m liking it so much better than Windows I wish I could just switch over. For now I need to recover my data from this piece of trash lacie. My terminal (found it), doesn’t know what mdadm is. Number 5. didn’t work, and number 14 obviously didn’t work since number 5 didn’t work. Where do I get mdadm and how do I install it?
Harry
September 3rd, 2009 at 04:38 AMLink to this comment
sudo: apt-get-install: command not found
Is the error message I get in the Terminal when I try number 5.
Harry
September 3rd, 2009 at 04:51 AMLink to this comment
sudo: apt-get-install: Turns out the – between get and install needs to go away. This works! Going to try again.
sudo apt-get install mdadm
Harry
September 3rd, 2009 at 06:05 AMLink to this comment
Whenever I try to save the mdadm after the ctrl X, and Y, I get ERROR writing, no such file or directory I can’t seem to get a break!
Harry
September 3rd, 2009 at 04:04 PMLink to this comment
OK I’ve got past all that. MDADM wasn’t installing. Kept coming up with this email config page and I was totally lost. Not sure how I got past it. Now I simply cannot get it to work. The best I’ve been able to get is this message mdadm: device /dev/md0 already active – cannot assemble it
Harry
September 3rd, 2009 at 04:08 PMLink to this comment
I unplugged the USB and I still get the same message so perhaps my config file is still messed up. I found having UUID on the 3rd line brought up a whole other set of errors so I put it on the ARRAY line. Now this.
Harry
September 3rd, 2009 at 04:13 PMLink to this comment
I change the config file to say /dev/md1 instead of 0 and assembled md1 and it worked!
Harry
September 3rd, 2009 at 06:26 PMLink to this comment
Here is what worked for me.
mdadm.conf (cleared out all the help statements first) then typed…
DEVICE /dev/sdd2 /dev/sde
ARRAY /dev/md1 level=linear num-devices=2 UUID=163218f3:299b9ccc:3c666c38:6053
sdd was originally sda and sde was originally sdb in the LaCie file.
No matter what I did, it wouldn’t work until I changed md0 to md1. At that point the command sudo mdadm—assemble /dev/md1 did the trick.
Harry
September 3rd, 2009 at 06:28 PMLink to this comment
ARRAY /dev/md1 level=linear num-devices=2 UUID=163218f3:299b9ccc:3c666c38:6053 This line looks like 2 lines and was 2 lines on the LaCie file but it has to be 1 line in the new mdadm.conf file. Just put a space between “devices=2” and “UUID”
Harry
September 7th, 2009 at 04:19 PMLink to this comment
FYI I inspected the controller card from the EBD, and it has a blown/ leaked capacitor. This is undoubtedly caused by the failed PSU. Perhaps simple using a REAL PSU will prevent this for those that still use these drives.
dada
September 25th, 2009 at 07:14 AMLink to this comment
Thanks, it also works for “jbod” linear partition on a d-link dns-323 . Great!
Snorri
October 3rd, 2009 at 05:18 PMLink to this comment
your welcome :)
Joe
November 18th, 2009 at 09:45 PMLink to this comment
Thank you for this great information. I followed the instructions without any errors but I still cannot access the data on the drive and get the following:
administrator@ubuntu:~$ sudo fdisk -l
Disk /dev/sda: 80.0 GB, 80026361856 bytes 240 heads, 63 sectors/track, 10337 cylinders Units = cylinders of 15120 * 512 = 7741440 bytes Disk identifier: 0×72722ea5 /dev/sda1 * 1 10337 78147688+ 7 HPFS/NTFS
Disk /dev/sdb: 500.1 GB, 500107862016 bytes 255 heads, 63 sectors/track, 60801 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Disk identifier: 0xfb21f290 /dev/sdb1 1 125 1004031 5 Extended /dev/sdb2 126 60801 487379970 83 Linux /dev/sdb5 1 16 128457 82 Linux swap / Solaris /dev/sdb6 17 17 8001 83 Linux /dev/sdb7 18 18 8001 83 Linux /dev/sdb8 19 34 128488+ 83 Linux /dev/sdb9 35 125 730926 83 Linux
Disk /dev/sdc: 500.1 GB, 500107862016 bytes 255 heads, 63 sectors/track, 60801 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Disk identifier: 0×00000000
Disk /dev/sdc doesn’t contain a valid partition table
Disk /dev/md0: 999.2 GB, 999184924672 bytes 2 heads, 4 sectors/track, 243941632 cylinders Units = cylinders of 8 * 512 = 4096 bytes Disk identifier: 0×00000000
Disk /dev/md0 doesn’t contain a valid partition table
Any suggestions on how to fix this without losing the original data?
Thanks Joe
Dave
November 23rd, 2009 at 12:01 AMLink to this comment
Thanks, very helpful. I had to use the build command and wasn’t able to mount /dev/md0 … but after rebooting ubuntu (VM under OS/x) and then powering up the drives – a file browser window for the restored raid appeared. A little finagling with file sharing and I was able to copy all my files! Why couldn’t LaCie provide restore tools? You know they have them….
Again, many thanks.
GD
December 5th, 2009 at 09:56 PMLink to this comment
Great post and closest I have gotten to recover data from my Lacie.
I have Ubuntu 9.x installed and went through this step by step. I’m stuck on step 10. I have no 130mb mount. One change though, I have both Lacie drives (2×500) plugged into the internal sata ports (0 & 1 with my Ubuntu on port 3). Ubuntu tells me that there is a problem with one of the Lacie 500Gb (many bad sectors). This is the result of fdisk:
Disk /dev/sda: 500.1 GB, 500107862016 bytes 255 heads, 63 sectors/track, 60801 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Disk identifier: 0×00000000
Disk /dev/sda doesn’t contain a valid partition table
WARNING: GPT (GUID Partition Table) detected on ’/dev/sdc’! The util fdisk doesn’t support GPT. Use GNU Parted.
Disk /dev/sdc: 500.1 GB, 500107862016 bytes 255 heads, 63 sectors/track, 60801 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Disk identifier: 0×00000000 /dev/sdc1 1 121603 976773119+ ee GPT
Disk /dev/sdb: 500.1 GB, 500107862016 bytes 255 heads, 63 sectors/track, 60801 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Disk identifier: 0×0000a7bf /dev/sdb1 * 1 59358 476793103+ 83 Linux /dev/sdb2 59359 60801 11590897+ 5 Extended /dev/sdb5 59359 60801 11590866 82 Linux swap / Solaris
Step 11 works or at least I can do it but with no step 10, 11 is no good.
A little more background here explaining the condition of my drives… before this, I tried Data Rescue 3 which scanned and attempted to recover my data taking 7 full days. It recovered the data, I can see the files but almost all of the files are invalid or corrupt. I tried Disk Warrior and same thing. so called recovered files are corrupt and invalid. I also tried cloning the Lacie using Data Rescue to a 1Tb drive and then scan/recover, which takes only 24 hours but same results. The files are either invalid or corrupt. The file sizes are correct and the total data recovered (about 300Gb) is correct but un-usable files.
Is my drive so damaged that nothing will work?
Is there anything I can do using this method to get my data back? Any other tools?
Thanks all!
Snorri
December 18th, 2009 at 12:03 AMLink to this comment
Hey guys,
I’m no linux expert, just managed to figure this stuff out by being stubborn! :P
I recommend you get more help over at ubuntuforums.org for example. Those guys are probably more help than I am and are more familiar with problems like bad partitions.
Don’t give up tho… :)
Also avoid running anything that writes on your disks or tries to repair something, that will probably ruin your data.
Chris
January 1st, 2010 at 09:57 PMLink to this comment
Amazing – thank you SO much for this tutorial. You saved my bacon!
Chris
January 10th, 2010 at 01:21 AMLink to this comment
Hello, Snorri. Thanks for this tutorial – it looks promising. However I’ve run into trouble – only 1 of the disks is showing when I run fdisk (shows 500GB and it’s the one without partition). So, I have a couple of questions
- does it matter in what order they are inserted into the enclosure? - or, does this mean the other disk is dead and unrecoverable?
I’m using Ubuntu 8 in a virtual machine (Parallels, on a Mac)
Thanks
Snorri
January 12th, 2010 at 01:07 PMLink to this comment
hey Chris, check your email.
S5 Sonny
February 25th, 2010 at 05:52 PMLink to this comment
Hey I seem to be having a problem where both drives have a doesnt contain a valid partition table error.
I went ahead and followed your steps anyway. When i run the mdadm—assemble /dev/md0 I get a not enough parts in an array error Is there anything else I can do?
Antonio Correnti
March 6th, 2010 at 04:56 PMLink to this comment
Hi, Thank you for the guide! I have one question: Do I need a box that supports jbod or a normal Dual sata should be ok? And if I get a jbod one, should not work out of the box? Why do we need such settings? Thanks for the answers!
Antonio
Ryan
March 19th, 2010 at 12:51 AMLink to this comment
I was able to pull the drive enclosure apart, plug in PSU (which was not providing enough power but was enough to power the circuit etc) and plug in two sata power connectors from a PC. Turned it all, used the IP Configurator and was able to get all the data off. Easier than using Ubuntu live CDs etc.
snorri
March 25th, 2010 at 10:53 AMLink to this comment
Sonny: can’t help you there mate. Disks might be damaged.
Antonio: No, just a normal enclosure. It’s a software JBOD so Ubuntu should take care of that for you.
Antonio Correnti
April 8th, 2010 at 01:52 PMLink to this comment
Hi, I did it with a 35$ usb dual sata dock and it worked! So I get no chance to get it work on mac os ? Thank you!
Manfredo
July 30th, 2010 at 04:42 AMLink to this comment
Duuude! I just wanted to thank you for this post, the instructions were pretty straight forward, for the past two weeks I’ve tried every f..ng recovery software out there, and none worked, Last night when I was ready to give up, I came across your site, I was pretty sure it would work so I downloaded Ubuntu destop, booted up on LTS (live CD without install) follow your steps, and VOILA! I was staring at my files amazed on how simple it was… Of course I’m copying 860GB of data to a brand new USB 1.5TB HD, it has copied almost half of that without a problem, and I’m sure it will be able to copy all data, since the RAID is mounted as it was on Lacie’s enclosure. So thanks for showing me “the light” hehehe :)