Data recovery nightmare

[mslaterhk]

Ok so I'm usually a pretty level-headed guy but this has me sweating - hope someone can help...

QUESTION:

If ReadyNas has somehow over-written (MBR?) root folder structure of logical volume, and my NAS is dead, can I use some restore program to rebuild file structure or at least recover all the data on the drive (via external USB to SATA connector)?

Any help much appreciated!

BACKGROUND:

Months ago ago I posted about warning messages regarding the ReadyNas Ultra Duo sending repeat messages about unexpected voltage levels (see here: https://community.netgear.com/t5/Using-your-ReadyNAS/V-12-power-is-out-of-normal-range/m-p/945819#M7...

After a lot of searching I did not find any resolutions, and it seems this is something that may happen to a nearly 5 year old NAS, so I sucked it up, bought an external 2TB drive and scheduled daily backup of the NAS to this drive in case the whole lot goes.

Over the weekend the ReadyNas started significantly failing, where the data volume would dissapear from my Mac often.  You could hear a solenoid or something click on and off in the NAS unit when this happened.

I went into admin mode and found that somehow this had triggered one of my 2TB drives to start re-synching with the other, showing 3% sync completed.

Then the click occurred again, logical volume dissapeared, then came up 30 seconds later, this time, no more syncing.  But all data appeared to be gone!  I switched NAS off at this point for good.

On attempting restore of data from the external backup disk found that the external drive attached to NAS has failed also!  All HW tests to the drive fail - it won't spin up at all.

At present I think I have:

- Dead backup disk

- 1 disk which has been in progress of re-syncing with good data for 3% or more.

- 1 disk which should be ok.

I have put both of the NAS disks in an external enclosure and spun up an Ubuntu VM and attempted reading them (after researching and mounting them as Linux RAID disks).

The disk which was re-syncing shows invalid.

The "good" disk comes up with a Linux root drive - a lot of standard folders - var, usr etc.

it has a c folder, and a file at root which is named the logical name of my now dead NAS volume.

However when cd into c folder there is nothing there.

[mdgm-ntgr · Accepted solution]

What you are looking at is the 4GB OS partition.

You will need to install mdadm and lvm2 if they are not already on your system

# apt-get install mdadm
# apt-get install lvm2

If you have not already you need to start the RAID

# mdadm --assemble --scan

We use LVM on legacy ReadyNAS so you need to do

# vgscan
# vgchange -a y

Then you should be able to mount the volume

# mount /dev/c/c /mnt

where /mnt is the mountpoint (could be anything, needs to be an empty directory).

On ubuntu you would need to put sudo at the start of these commands, I think.

[mdgm-ntgr · Accepted solution]

What you are looking at is the 4GB OS partition.

You will need to install mdadm and lvm2 if they are not already on your system

# apt-get install mdadm
# apt-get install lvm2

If you have not already you need to start the RAID

# mdadm --assemble --scan

We use LVM on legacy ReadyNAS so you need to do

# vgscan
# vgchange -a y

Then you should be able to mount the volume

# mount /dev/c/c /mnt

where /mnt is the mountpoint (could be anything, needs to be an empty directory).

On ubuntu you would need to put sudo at the start of these commands, I think.

[mslaterhk]

Many thanks for the reply!
I should have mentioned I have already installed MDADM.
I can read the partition properly, but the c folder is empty.

By doing the vgscan and vgchange and then the mount will this allow me to view contents of c folder?

[mslaterhk]

ok just read the first part about partitions. I'll respnd shortly with list of my partitions and contents.

[mdgm-ntgr]

---

@mslaterhk wrote:



By doing the vgscan and vgchange and then the mount will this allow me to view contents of c folder?

---

All going well, yes.

You may wish to do a read-only mount e.g.

# mount -o ro /dev/c/c /mnt

---

@mslaterhk wrote:
 but the c folder is empty.

---

It is a mount point. Mount points are empty directories.

[mslaterhk]

Firstly, a really big thank you for your replies.

I've mounted the volume now and data integrity is good.  Your commands work perfectly!

So this thread can be marked as solved now.

One more thing as I'm not a Linux command line guru:

I'm trying to copy my files in the gui but keep coming up with "cannot copy .AppleDouble" folder.  There are literally thousands of these.  I'm trying to copy to new NAS network share.

should I mount network share in cli and use cp and ignore errors, or is there a way in the gui to ignore the errors when copying?

Many thanks!

[mslaterhk]

And as an FYI if anyone else comes across this and tries to do the same thing:  for some reason when using a virtual Ubuntu and attempting to copy files from the mounted disk, Ubuntu will crash if you try to copy the files directly from mounted disk to another USB mounted drive.

I've had to copy from RAID disk to local folder, and then copy from local folder to actual destination.

Don't know why, but copying to local first seems to work.  Hope it helps someone.