Help recovering data from SC101 drive

[Jack™]

Originally Posted by bigmccracker

Sorry, I had upgraded firmware and now get a "Corrupt superblock" error with SFSExtract. Z-scanplus shows 1 MicroSAN, but no partitions.

Post a 'z-scan listall' and 'z-scan getpart' for each found partition. See my sticky How to post 'listall' and 'getpart' information - the complete guide for more information.

Originally Posted by bigmccracker

The disk is found in SCM, but I get the message that it will be formatted and all data lost when I try to re-attach it.

This formatting question you get is from the Windows explorer not the SCM. First try this: SC101/SC101T - How to fix un-allocated partitions?

[bigmccracker]

Thanks again for the advice, Jack. The partitioning you mentioned won't work, because Disk Mgmt doesn't see an unallocated partition. I'll post the listall and getpart results in a moment, but I'll describe my problem a little further.

I had originally set up the SC101 about a year ago in a mirrored configuration (2 320GB drives). A few weeks ago the drive disappeared. (I'd had no problems and hadn't had to touch it in all that time.) When I opened SCM to see what was happening, I got a "Broken Mirror" error. When I tried to "Manage Broken Mirror", the response was something like "no mirrors found."

I searched through these forums and determined that I should try to upgrade the software and firmware and then use some of the tools available. Which I did, and both seemed to work fine, but did not result in me being able to re-attach the drive. "No drives available to attach" is all I get as a response. Here are the listall and getpart results you wanted:

C:\Program Files\NETGEAR\NETGEAR Storage Central Manager Utility>z-scanplus list
all
++++++++++++++++++++++++++++++++++++++++++++++++++ +++++++++++++++++++++++++++++
Network Adapter : VIA Rhine II Fast Ethernet Adapter
IP Address : 172.16.1.64
Subnet Mask : 255.255.255.0
Discovering uSAN with this network adapter, please wait...
Finding roots...
Found root IP = 172.16.1.94
Found root IP = 172.16.1.81
root 172.16.1.94 "" has 1 partitions
Market class=0 64, Manuf code=9 64 241, Part num=0 0 101
root 172.16.1.81 "" has 0 partitions
Market class=0 64, Manuf code=9 64 241, Part num=0 0 101
================================================== =============================
VERSION : 4.23.0 ROOT IP ADDR : 172.16.1.94
MANUFACTURING CODE: 606449 PART NUMBER : 101
TOTAL(MB) : 305240 # PARTITIONS : 1
FREE (MB) : 88
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
PARTITION LABEL IP ADDR SIZE (MB)

853A5092-5849-11DD-BE4A-0016178E63EE.m1 Cracker6 172.16.1.95 305152
================================================== =============================
VERSION : 4.23.0 ROOT IP ADDR : 172.16.1.81
MANUFACTURING CODE: 606449 PART NUMBER : 101
TOTAL(MB) : 305240 # PARTITIONS : 0
FREE (MB) : 305240
================================================== =============================
2 MicroSANs found within subnet 172.16.1.255

C:\Program Files\NETGEAR\NETGEAR Storage Central Manager Utility>z-scanplus getp
art 172.16.1.94 172.16.1.95 password
================================================== ============================
USER NAME : Cracker6
UNIQUE NAME : 853A5092-5849-11DD-BE4A-0016178E63EE.m1
READ TOKEN : cracker1
WRITE TOKEN : cracker1
ATTRIBUTE : 00 00 00 00 00 00 FF 00 02 00 01 00 00 00 02 00
: 00 11 22 33 44 55 66 77 88 99 AA BB CC DD EE key
r r r r r r r MC MC VM VM BS r r Str/Wr/Sp/Sh/Ow
================================================== ============================
MirrorCount=2 ValidMirrorCount=1 StripeBlockSize=0 (0 KBytes)
Owned=0 Shared=1 Spanned=0 Written=0 Striped=0

C:\Program Files\NETGEAR\NETGEAR Storage Central Manager Utility>

I pulled the broken drive and attached it through an external enclosure, to get a Windows drive letter, and ran SFSExtract. I got "Corrupt superblock" messages on both SINGLE and THOROUGH and nothing on WDISK. (I can post those results, if you want.)

I've attached the drive through the external enclosure to another computer and ran Linux DD to pull the image, but when I wrote the image to another drive and put it into the SC101, the SC101 didn't even identify it. SFSExtract when connected through the external enclosure gave the same response as with the original drive. (I've since re-formatted it to try again, but haven't tried again.)

I realize that I'm probably making things worse every time I try anything, but I'm kind of desperate to get the data from the drive.

[Jack™]

Originally Posted by bigmccracker

I realize that I'm probably making things worse every time I try anything, but I'm kind of desperate to get the data from the drive.

So hold your horses and don't try any thing you don't know what it will do (if you care for the data, that is)............

1. You have now the latest firmware 4.23.0, but nobody knows what firmware you had when you first encountered the problem.
2. One disk with root IP 172.16.1.81 has the superblock problem as there is no partition information to be found. Take out this disk from the SC101 box.
3. Try to attach again.
   1. If fails, perform a long (>45 sec.) reset and try to attach again.
   2. If this also fails, try to attach from a PC that was never used for working with the SC101 and you therefor need to do a first install of the software 3.0.6. Download software and drivers from here: SCM Utility 3.0.6 .
   3. You could also remove the software and drivers from the PC you are now using, clean the registry and then after 2x rebooting, reinstall the drivers and software.

FYI, the SCM does not repair mirrors and if it does something, then it splits broken mirrors in 2 regular partitions so you'll have access to the data on both parts).
Never put the disks in an external USB enclosure or install the disk in a PC and start twiddling ...... (again, if you want to keep the data....)
The SFSExtract utility works via the SC101 enclosure and only after attaching a drive, which will assign a drive letter to the partition.

I won't tell you now what to do if any of steps 3.1, 3.2 or 3.3 above fail.
Let's first get to that bridge before crossing it....

[bigmccracker]

I'm not quite sure what you mean by "perform a long (>45 sec.) reset."

Do you just mean holding the reset button for more than 45 seconds? Do I hold the button down, power on the machine and keep holding it for more than 45 seconds?

You're not talking about a "factory reset", are you?

[Jack™]

No, I don't mean a factory reset or Telnet reset!
The reset you describe is also known as a 'hard-hard reset' to get "frozen" routers responsive again.


A SC101/SC101T long (>45 sec.) reset is performed as follows:
While the unit is powered up and the disks are inserted, you push the reset button for at least 45 seconds and wait until unit is again online.

This should reset the unit's password to the default 'password' (not the write/read passwords of individual drives) and should detach all drives.

Some users reported that detaching occurred only after a prolonged reset of longer than 60 seconds while others reported that it took more than one reset cycle to achieve the required result.

This reset does not erase any user data on the disk(s)!

[bigmccracker]

I haven't had much time to work on this, but the few times I've tried, I've had no luck getting the system to reset. I've been testing by trying to change the admin password and using "password" as the current password, but I get a message that the current password is incorrect.

Within the next day or two, I'll try connecting from a different computer and let you know the results.

[bigmccracker]

I have not been successful in resetting the admin password, so I don't think the reset has worked.

I have been able to attach the "working" drive (172.16.1.94) to a new computer with a new SCM 3.0.6.

There is something wrong with the mirror. Perhaps I had set things up wrong in the first place. I moved about a year ago and had to set everything up all over again. The data on the working disk is a year old. There doesn't seem to be anything new since the move. I'm hoping that the last year's worth of data is on the "broken" drive (172.16.1.81).

In the end, covering steps 3.1 and 3.2, I still don't have access to the broken drive. (Unless there's some tips to get the reset to work.)

[Jack™]

I don't know if the reset did or did not work. But at least, you were able to attach one part of the broken mirror.
The other part of the mirror is located on a disk which superblock is apparently damaged.

If you think there could be newer data on that disk, then the following procedure(s) can be done:

1. Take out the HDD that could be attached and check if you can attach and access the data on the other damaged HDD. If this is possible, copy all available files to another media. If this fails, insert the disk in the other slot and try again.
2. If the data remains inaccessible, then, as long as you obtain a driveletter under Windows, use the SFSExtract data recovery tool. See: SFSExtract Syntax .
3. Without obtaining a driveletter, you may try the Wrapper procedure (Dogbert's post #28) or the Linux disk image copy procedure (see Rodyz's post #16) .

[bigmccracker]

I don't understand "the Wrapper procedure (Dogbert's post #28)" information. There don't seem to be any procedure-type instructions. Do I need to call Netgear support for help on that one?

I'm going to try the "Linux disk image copy procedure (see Rodyz's post #16)" again and see if I get any results.

[k6_user]

bigmccracker

In post #18 you asked "Did this method prompt to re-format and lose all data? I'm not sure I'm willing to risk losing everything."

Yes it did, however if you read my procedure in post #12, step3, you'll see that I had first used the DD command to make a copy of the disk, and since I was using the copied disk in the SC101 I wasn't really risking losing the original data. Besides, as Jack said, using the "wrapper" method as I did in step 4, in which you tell the SCM to use the whole disk as one partition, doesn't really wipeout the data.

I see in post #22 that you did copy the disk, but then said "the SC101 didn't even identify it" If by that you meant that the SC101 didn't even see the disk as a new unformatted disk, then perhaps the disk is defective or not compatable with the SC101 (Remember to set the jumper to CS). If you meant that the SC101 didn't see it as an "attachable" disk, thats probably due to the corrupt super block and you'll have to use the "wrapper" method to access the disk in the SC101.

In post #29, I share your frustration. I'm no expert, but as I understand it the "official" wrapper procedure requires Netgear support to perform it via remote access to your computer. In Dogbert's post #28 he says it works in a similar fashion. It seems that you can skip step one because the SC101 doesn't see any partition to delete. Step two is to just use the SCM as normal to create a new partition that uses the entire disk - as I did, and it worked for me, Just be sure to us a copy of your original disk - just in case.

Good luck. I hope you're successful using Rodyz's procedure (post #16). The main difference between his and my procedure (post #12) is that I had to do the "wrapper" method in step 4 in order to get the SC101 to see my drive. Then I was able to use SFSExtract.

Sorry for the delayed reply; I've been busy reconfiguring my network/backup scenario. I'm convinced that having two drives in an SC101 creates a heat problem which puts too much stress on the hard drives. I lost three drives (2 Maxtors and 1 Seagate) in about one year's time. I picked up another SC101 (ther're going for about $30 incl shipping on eBay) and now have two SC101's with one drive in each (runs a LOT cooler). Since then, my son gave me his old Linksys router which allows sharing a USB drive on the LAN; so now I have triple protection. Do I sound paranoid?