Whoa... REALLY bad situation... Need Help

[apexdemon]

OK... my situation has gone from bad to worse and I need all the help I can get. Appreciate if all you gurus out there can give this a read...

Let me try and summarise the key parts of what has happened with my N2100 and its two Samsung HD501LJ (500GB) disks in RAID 1:

- A few days ago, I got a degraded RAID 1 warning. I subsequently tried a raid rebuild through the web admin interface which got stuck at 99.9% (for like 12 hours) the first time. Finally I used the Shutdown function from the web admin and tried again.

- The 2nd time I tried the rebuild, it automatically retried from scratch after reaching like 60+% and eventually kept retrying from a smaller and smaller percentage of completion each time. This freaked me out and realizing that my bad disk was getting worse, I shut it down from the web admin, and went out to order the same disk for replacement (which may arrive in about a week).

- In the meantime, since I could still mount the shares from Windows, I decided to copy out whatever data I could over Ethernet but after a big overnight transfer, I woke up to find that the process had failed midway and the NAS had disconnected itself. This time however, I was having intermittent problems logging back into the web admin interface and couldn't shutdown from the admin, and couldn't shutdown by pressing the power button on the fascia either. No choice... shutdown by disconnecting power.

- I then went out again and bought an external sata disk enclosure, popped the "good" disk in there, downloaded and installed Ext2 IFS
(http://www.fs-driver.org/index.html) and managed to mount the volume in XP but couldn't access it. Downloaded and ran mountdiag.exe as suggested by the Ext2 IFS FAQ and it reported that the disk was not cleanly unmounted and therefore could not be accessed.

- So I put the disk back into the NAS and rebooted, hoping that I would be able to attempt a proper shutdown to resolve the clean dismount but this time the web admin interface is not showing up at all. Essentially it is not found and I think the web server aspect of the NAS is not running at all. Pinging the IP address works, but no web admin interface to speak of.

So now it looks like I'm buggered. I'm a casual user of the NAS so I haven't installed any SSH or other modules in the past, I don't even know for sure what the firmware version is (may not be the latest but should be fairly recent, at most 1 or 2 generations from latest). I'm ready to ditch the NAS completely, all I want is my data rescued. I've got photos and home videos of my kids, our holidays and some work backups that are really irreplaceable and which I naively thought would be safe with a Raid 1 setup.

Pls help with ideas on how I can rescue the data. THANKS!

[dbridges]

Ok tackling the data thing.

For anyone reading this raid usually degrades for a reason. At that point your first priority is moving your data to another secure location. Then you try and rebuild. Raid 1 is only as good as both disks. If one disk is faulty then your data is not secure.

Are you sure that you correctly identified the good disk?

Did you try both disks.

No need for an external enclosure if you can put them in your machine directly.

If both are still faulty then try system rescue cd mentioned in the wiki. It worked for me.

As for the web gui.

Try https instead of just http. If that works turn http back on through the web gui.

Otherwise try the factory reset (download the manual from thecus, the instructions are in the back).

SSH may not be useful because your disks are in a mess so don't worry too much about not having it.

While you've got your disks out you can add a root share (See the wiki) and use that to get access to your configuration files. You'll be able to restore the web gui that way if you need to.

You may find that you need to replace both disks.

[apexdemon]

dbridges,

Thank you so much for the advice! I had given up on this forum after eager daily checks and finding no response. Bless you!

I will be giving this a try this weekend when I have time and will report back here in the hope that whatever the outcome, someone else can learn from it.

[apexdemon]

So here's an update on my situation...

Finally managed to figure out the creation of a SystemRescueCD disk (although I skipped the checksum portion... couldn't figure it out and luckily it turned out I had a clean download so didn't need it). Booted from the CD and used GParted to "clean up" the better disk of the 2. Remounted it in windows using Ext2 IFS and could actually see my NAS files! I was ecstatic! Altho I suspected not all was entirely well as some folders took a while to open etc. For some inexplicable reason, my PC started to emit a two-toned siren alert (from the motherboard's piezo speaker, not the soundcard/external speakers). At this point I had no idea what the alert tone was so I shut down windows (properly). Thinking there was something amiss with the linux fix of the disk filesystem and that was somehow causing windows to scream blue murder, I rebooted into linux with the cd to check the disk again, and now even linux was emitting the same motherboard siren. Shutdown (properly). Research from another computer indicated that this could be an overheating issue so I spent the next hour opening up the pc case and directing airflow at it. Rebooted into windows and the "rescued" disk could no longer be read. Rebooted into linux to attempt a refix but this time GParted just disappeared from screen halfway thru the process. I figure this was a crash. Repeated the process with the known bad disk (the other disk from the raid 1 array) and things were pretty much the same, just worse.

So bottom line is, I was not able to rescue my data. Maybe if I had started transfering my data during that one fleeting moment when I could view my data in windows (albeit in a flaky manner) I might have saved some of it. But because the overheating siren came on, and my ignorant panic that set in, I now have nothing, nada, zilch, zip.

Moral of the story? Turns out there are quite a few...

1. Don't trust your Raid 1. Two disks CAN fail at pretty much the same time. Get to know the procedures for recovery before settling into the delusional bliss of having your data safe forever.

2. Don't get a NAS with Ext2 filesystem unless you have a linux box you can readily plug the disk into should one disk fail.

3. Don't do anything else if there is a raid degrade warning... get the data out first before anything else. If you really must have your data safe, heck, buy TWO NAS arrays or run Raid 5 or something.

4. Always thank the heavens for people like dbridges who will try to come to your rescue.


If anybody else has any ideas on how to rescue the data from my (obviously) very-screwed-up disks, please share.

THANKS!

[wengi]

Hi,

simply some additions to your conclusions:

I have seen two drives failing at the same time several times. The drives were allways bought at the same time from the same dealer, so it was the same lot of drives (simialer serial numbers).
This happend here in a RAID5 with 7 drives and one hot spare.
The array was running for over six month when the first drive crashed. The controller started to copy the data to the 8th drive (hot spare). But the next drive collapsed within 30 minutes.

over 1 TB of productive data nearly lost....

Conclusions:

- Raid5 does not help if two drives fail.
- You have to act as soon as possible when dricves fail.
- The best is to backup your data on a minimum of two devices.

I feel with you and hope there may be any way to save your data.
You could e.g. try to change the board of the two dricves. That was how i rescued the data...

wengi

[dbridges]

Apologies. I dont get the chance to get on as often as I used to.

Sorry to hear that it hasn't worked out.

You should have been able to do the move from within system rescue. No need to go back to windows.

The buzzer you experienced was quite strange. You suspect heat but for me heat has usually just powered down the machine...

As a side note: I recently has some disk problems with a faulty HDD in my PC and spend a lot of time diagnosing it to be a cable issue.

There were some additional problems that I had to overcome though. The first was that the NTFS filesystem had been corrupted which I repaired using debian (of all things).

The second was that the nvidia ide controller had downgraded the drive to pio (or lower) and refused to automatically "re-grade" it to a better condition after the cable was replaced. Windows just couldn't handle it. I had to manually upgrade it's status.

Moral of the story. There are lots of factors in play and the problem may not be the one that you expect.

[subsubone]

www.data-recovery-software.net/Linux_Recovery.shtml

Had a near same problem as yourself. The above saved my life and sanity, and it's FREE!

Hope it helps you as it has helped me.

Cheers,

Leigh.