Troubleshooting Help - Dead Volume After Loose Cable Cycling Power

[kevroy316]

Today, I was installing a new device near my ReadyNAS 104 (4x 4TB), and I noticed as I was finishing up that the power cable has apparently stopped staying in reliably and had been fluttering in and out of the socket (cycling the power rapidly). I went ahead and secured everything and booted up. It booted ok, but all of the volumes were gone.

I looked into it a little more and the data-0 volume went from redundant to degraded and then degraded to dead in the course of a couple minutes as a result of the power cycling. I've never had any problems until this, but I honestly have no idea what to do now.

I have backups, but they're not entirely up to date and due to the volume, they would take a LONG time to restore. I'm inclined to go find a new product that doesn't have such simple issues (I've also not been happy with its performance in genral), but in the mean time, I'd like to do whatever I can to get things back up and running.

I've read through several forum posts with people in similar situations, but power cycling, leaving it off for a little while, and looking for ATA errors hasn't revealed anything particularly helpful. 

The current error just says "Remove inactive volumes to use the disk. Disk # 1, 2, 3, 4."

The only option I feel like I have is a factory reset or reformat followed by a restore from backups. As this would take a couple weeks to finish, I'd like to try a few more things before I resolve to that.

Any advice would be very helpful!

[Sandshark]

The existence of a data-0 volume is usually an indication that one drive got out of sync and then the NAS tried to mount it separately.  Since there cannot be two volumes named "data", it calls the second one "data-0".

Now, if you are lucky, three of the drives are still in sync and form a good, though non-redundant, volume.  The trick is figuring out which three.

The Volume tab may give you some hints.  Hover the mouse over each drive and one of the things that will show in the pop-up is what volume(s) the drive is a part of.  If three are a part of one volume (may be "data" or "data-0"), then those are the ones to try first.  Alternately, you can just try all combinations of three.  When doing so, it is best to use the boot menu and boot up in read-only mode to discover which combination works, then do a normal boot with that combination.  If it boots in normal mode with three drives, you can hot-insert the 4th, have the unit format it, and then it will re-sync.  Removing all partitions before insertion can sometimes make that go easier.

If you are unlucky and two drives were affected, then you are looking at a data recovery (expensive) situation or just kissing your non backed up data goodbye and doing a factory reset and backup restore.

[kevroy316]

Thank you for the reply. After fiddling around in the terminal for a bit via ssh, I'm fairly certain things are out of sync. I'm really a complete novice at how RAID works, but when I run mdadm --assemble --scan --verbose, I see several /dev/sd[abcd] values that have "No super block found... (Expected magic ######, got 0000000)" or similar. There are 4 sd[abcd]3 elements that are labelled 'slot 0' through 'slot 3' and contain about the right amount of space (~3.5TB each), but I'm not sure what to do with them to make it work.

I'm fairly resolved to losing everything since the last complete backup (though I have no idea how to get CrashPlan to give me 3.5 TB of data back - their download speeds are not good), but I'd like to make sure I've exhausted my options.

My next step is to unplug all the drives, stick them in my desktop, and try using testdisk.

Definitely open to try other steps though. I'm not going to give that shot until tomorrow evening.