How to get data out of a Maxtor Shared Storage II that fails to boot?

Question

I've got a Maxtor Shared Storage II (RAID1 mode) which has developed some hardware failure, apparently: it fails to boot properly and is unreachable via network.

When powering it on, it keeps making clunking/chirping disk noise and then sort of resets itself (with a flash of orange light in the usually-green LEDs); it then repeats this as if stuck in a loop. In fact, even the power button does nothing now – the only way I can affect the device at all is to plug in or pull out the power cord!

^{(To be clear, I've come to regard this piece of garbage (which cost about 460 €) as my worst tech purchase ever. Even before this failure I had encountered many annoyances about the drive: 1) the software to manage it is rather crappy; 2) it is way noisier that what this type of device should be; 3) when your Mac comes out of sleep, Maxtor's "EasyManage" cannot re-mount the drive automatically.)}

Anyway, the question at hand is how to get my data out of it? As a very concrete first step, is there a way to open this thing without breaking the plastic casing into pieces? It is far from obvious to me how to get beyond this stage:

It opens a little from one end but not from the other:

If I somehow got the disks out, I could try mounting the disk(s) on one of the Macs or Linux boxes I have available (although I don't know yet if I'd need some adapters for that).

(NB: for the purposes of this question, never mind any warranty or replacement issues – that's secondary to recovering the data.)

Answer 1

Opening the case

Pull the gray side panels slightly apart at the front. Then lift the back of the dark "wrapper" piece (the one on which you loosened the two screws) and slide it forward. Do this both at the top and the bottom, then the wrapper piece will come off.

For the not-so-faint-of-heart: you might just as well rip the wrapper piece off – this will break all the tabs, but if you do decide to piece your MSS II back together, the two screws of the rear panel will still hold it in place.

You should now have exposed two screws on the front panel. Loosen them, and the two pieces of the metal case will slide apart.

You can then take out the disks. They are plain SATA disks, which you can install in a standard PC or USB enclosure.

Accessing partitions

Note: the following goes for a single-disk MSS II. YMMV on dual-disk models, use at your own risk.

To my knowledge, the dual-disk models are set up as RAID 1 (mirrored disks), thus data format should be the same on both models.

Partition layout is as follows:

• sdx1: System image, around 250 MB.
• sdx2: Copy of sda1
• sdx3: Swap partition, around 250 MB.
• sdx4: Extended partition, filling all remaining space:
  • sdx5: System configuration, around 500 MB.
  • sdx6: Data partition. Each share corresponds to one directory.

The 500 GB model I originally had was formatted with a standard MBR partition table. As the drive began to fail, I have successfully transferred its contents to a 300 GB drive I had lying around, and later upgraded it with a 1 TB drive.

System images (sda1/sda2) can be found on the net. If you decide to continue using your MSS, I strongly recommend keeping an image of it somewhere so you can easily restore it – the MSS II is notorious for failing to come back up after an unclean shutdown (such as in the case of a power outage).

Data can be easily read off the sdx6 partition after mounting it on a Linux machine. You might need root for some of the files. If you're not planning to resurrect the MSS II but just want to get your files out, you can just do sudo chown -R to get ownership of all shared files.

I had an issue with Linux being unable to access certain files with accented characters in them. Some of them were restored from backup – there is a possibility that the MSS messes something up.

RAID?

If you have a two-disk model, there are a few extra things to consider.

First, if the disk is making strange noises during power-up, it is possible that one of them has failed. Since you have RAID, there is a good chance only one of your disk has died and the other is still alive. Find out which one is still good, and use that for your recovery procedure.

After you have restored one of the two disks, you need to fix the RAID. (Obviously, if you had a hardware failure on one of the disks, you will need to get a replacement.)

First of all, create an image of at least sdx1 and sdx5 of the good disk, as well as a backup of the contents of sdx6 – so you can always go back in case something goes wrong.

Then insert the good disk into your MSS II and make sure the bad one is disconnected. Start the MSS and make sure everything is OK. If so, shut it down cleanly.

You can now try the following:

• Plug the bad disk back in, power up the MSS II and try to re-sync the RAID with onboard mechanisms.
• Connect the bad disk to a PC, wipe the partition table, and place it into the MSS II (pretending a disk has failed and you're putting in a new one). Either the MSS II will recognize the situation and restore the RAID automatically, or you need to start this manually.
• Connect both the good and the bad disk to a PC, then copy the good disk to the bad one using dd.

Should you really resurrect your MSS II?

I'll leave that up to everyone to decide for themselves. I had to go through the recovery procedure I described here a few times, and found the MSS II notoriously unreliable as far as recovery from a power failure or otherwise unclean shutdown is concerned, on top of recovery being a pain.

Personally, after the last failure, I decided to just extract my data and replace the MSS II with a much simpler setup, consisting of a Raspberry Pi with a USB disk, Raspbian, Samba and Webmin. This gives me recovery options such as console access or simply being able to plug the disk into any Linux PC, apart from a much more extensible setup. It does require more tech knowledge than just setting up a MSS II, but if you manage to recover from a MSS II failure, you're likely to be savvy enough to set up your own solution.

Answer 2

I initially assumed that the daya on the drives would be in a standard format and was about to post "once you get the drives out, plug one of them into a Linux machine (or any other machine booted off a decent LiveCD) and it should detect what-ever filesystems are on there".

A quick search shows my first thought to be wrong though - apparently a standard boot sector and partition table is not present on the drives. See this page and the ones it links to for more detailed information including how to rectify the situation manually andmaybe get at the data. Be very careful when writing to the drive to try update the partition table and get the data off - I recommend backing up a full image of it to another drive if you have one big enough lying about (using compression you need not use a drive as bit as it is, unless the data is encrypted).

As you have it configured as RAID1 both drives should contain identical data, so if one has properly failed you can use the other to try read your data. Of course if the device has gone propely mental and damaged that data, you might find it to be equally damaged on both drives...