Data Loss after One Drive Failure

[32Domi]

Hi everybody,

 

I have a question regarding parity and rebuilding a defective drive. I had assumingly a corrupted drive (or at least a drive failure according to unraid dashboard). So I got myself a new drive and swapped it in but now two days later I see, that at least one, I guess two shares are completely empty.

I would like to understand how this could happen, although I have a parity drive? Everything is back green now in Unraid, but, as I said, the entire content of one share is missing.

Please don't tell me, to have backups, I have a backup that I'm currently rolling back for the more important share, but I would rather like to understand, how this happened.
Is it possible, that parity data was written, while the drive was already partly corrupted? Is there (even if I have the most important things) a way to get anything of the data? To me it's even unclear, what exactly is wrong with that drive, it's a 9 months old WD Red (EFAX).

 

Best regards, maybe someone can help me out (at least to understand the problem).

32Domi

[itimpi]

57 minutes ago, 32Domi said:

Hi everybody,

 

I have a question regarding parity and rebuilding a defective drive. I had assumingly a corrupted drive (or at least a drive failure according to unraid dashboard). So I got myself a new drive and swapped it in but now two days later I see, that at least one, I guess two shares are completely empty.

I would like to understand how this could happen, although I have a parity drive? Everything is back green now in Unraid, but, as I said, the entire content of one share is missing.

Please don't tell me, to have backups, I have a backup that I'm currently rolling back for the more important share, but I would rather like to understand, how this happened.
Is it possible, that parity data was written, while the drive was already partly corrupted? Is there (even if I have the most important things) a way to get anything of the data? To me it's even unclear, what exactly is wrong with that drive, it's a 9 months old WD Red (EFAX).

 

Best regards, maybe someone can help me out (at least to understand the problem).

32Domi

You should post your system's diagnostics zip file (obtained via Tools->Diagnowtics) so we can get a better idea if the state of your system.

 

Some questions:

• Do you have any drives showing as unmountable in the Main tab?
• Didi you at any time while trying to recover the drive use the format option (hopefully not).
• Do you still have the original drive intact as it was at the point you removed it from the array?

[KoNeko]

If your parity the biggest Disk in the array ? i think if the parity is not the Biggest disk i can go wrong.

 

i might be wrong so feel free to correct me.

[trurl]

2 minutes ago, KoNeko said:

If your parity the biggest Disk in the array ? i think if the parity is not the Biggest disk i can go wrong.

 

i might be wrong so feel free to correct me.

This is not going to have any relevance. The only thing that happens when you try to use a data disk larger than parity is it won't let you start the array and it tells you why. 

[32Domi]

55 minutes ago, itimpi said:

You should post your system's diagnostics zip file (obtained via Tools->Diagnowtics) so we can get a better idea if the state of your system.

 

Some questions:

• Do you have any drives showing as unmountable in the Main tab?
• Didi you at any time while trying to recover the drive use the format option (hopefully not).
• Do you still have the original drive intact as it was at the point you removed it from the array?

Find attached the diagnostics zip

 

• I have not now, as mentioned earlier, I had, so I replaced the drive
• I'm not sure if I actually did format it while trying to rebuild the array with the old drive. I considered it but I'm not sure, if I really did (it was wednesday night at.. 1:30 or so).
• I have the drive next to me, I only tried to mount it with a xfs tool (without any succsess), but nothing else.

Thanks in advance

unraid-diagnostics-20200831-1412.zip

[JorgeB]

Disk4 looks freshly formatted, it's completely empty, try to mount the old disk with the UD plugin, if it doesn't mount please post diags after trying.

[32Domi]

Yes, Disk4 is the one, I swapped in.

 

I connected the drive, it's not showing up at all in unassigned devices.

[32Domi]

My bad, I didn't connect it properly. Now it showed up. Find the diagnostics attached again.

 

It's properly mounted and I even can See the files. Gonna do a super quick backup right now.

unraid-diagnostics-20200831-1448.zip

[trurl]

For future reference, you absolutely must not format a drive in the parity array if you want to keep its files. Format writes an empty filesystem to the disk. Unraid treats this write exactly as it does any other, by updating parity. So after formatting a drive in the parity array, parity agrees it has an empty filesystem, and rebuilding it will result in an empty filesystem.

 

When you have an unmountable disk the correct thing to do is attempt to repair its filesystem. Ask for help before making a mistake.

[32Domi]

Thanks a lot. This time, I was lucky enough, next time I'll first consult the forum before doing anything else!

 

Still I'm wondering, why the drive was shown defective in the array.

[trurl]

1 minute ago, 32Domi said:

why the drive was shown defective in the array.

Based on the rest of the thread I guess you mean why was the disk unmountable.

 

The SMART report for that Unassigned Disk doesn't have any bad SMART attributes, but it did record a number of errors during its lifetime. After you get the data from it you might run an extended SMART test on it.

 

So, that disk may have had issues, or maybe some other hardware issue, perhaps intermittent, or loss of power, prevented proper maintenance of the filesystem when the disk was written.

 

Basically this is the same sort of problem you have with another OS such as Windows, for example, when you run checkdisk on the disk and it finds problems that need fixing.