Hello everyone,

Few months ago I started having problems with my server, multiple drives started having read/write problem which caused disk to become Unmountable. I had to repair the filesystem using the -L option and then they would soon re-corrupt themselves, while I tried to fix the root issue one of my drive became disabled. The problem is that I now have 1 drive that's disabled and 1 drive that's Unmountable.

Having only 1 parity drive, I fear this will lead to data loss. I have since received replacement drives and I would like to salvage what's left of my data and get the server running back up again.

 

What order of things would be proper in this case? repair the Unmountable drive and then replace the offline drive and attempt to repair it?

 

Here is a log and a screenshot.

 

Many thanks.

 

unraid-diagnostics-20210227-1324.zip

Edited February 27, 20215 yr by sloob

You rebooted before getting those diagnostics so can't see what may have happened before that. Disk3 isn't giving SMART, check connections and post new diagnostics.

11 minutes ago, trurl said:

You rebooted before getting those diagnostics so can't see what may have happened before that. Disk3 isn't giving SMART, check connections and post new diagnostics.

Actually I shut down the system when I saw that disk 3 had gone offline to avoid further data loss, about a week ago. I just turned it back on today because I received new drives to install. I already checked all sata/power connection, and when I move the drive to a different sata port/power port it remains offline, indicating (I imagine) that the problem is with the drive. But I will try anyway

unraid-diagnostics-20210227-1403.zip

Edited February 27, 20215 yr by sloob

Disk3 doesn't even appear in those diagnostics.

 

Emulated disk3 is mounted though so you should be able to rebuild it and it shouldn't need repair. 

 

Except...

 

Looks like disk2 and disk5 are beginning to have problems also. How did you get to a point where you had multiple disks with SMART warnings?

 

Do you have Notifications setup to alert you immediately by email or other agent as soon as a problem is detected?

 

 

12 minutes ago, trurl said:

Disk3 doesn't even appear in those diagnostics.

 

Emulated disk3 is mounted though so you should be able to rebuild it and it shouldn't need repair. 

 

Except...

 

Looks like disk2 and disk5 are beginning to have problems also. How did you get to a point where you had multiple disks with SMART warnings?

 

Do you have Notifications setup to alert you immediately by email or other agent as soon as a problem is detected?

 

 

I don't have a notification setup, I'm not all that familiar with servers and drives to be honest, I thought if one drive fails I'll just shut down the system while I wait for a new drive to arrive. I didn't think the chances of 2 drives failing at the same time were high enough to justify changing them when they failed a smart report. Might have been foolish on my part or even careless but at the end of the day the server only stores tv shows and movies, would suck enormously to download them again if multiple drive failed but it wouldn't be the end of the world.

 

Should I start by repairing the unmountable drive and then rebuilding the new disk 3 or vice versa?

Edited February 27, 20215 yr by sloob

In order to reliably rebuild every bit of a failed disk, Unraid must be able to reliably read every bit of parity PLUS every bit of ALL other disks. Parity by itself can recover nothing.

 

So obviously you don't want to wait until a disk completely fails to consider if any disks have problems since problems with any other disks that might not have completely failed yet can compromise any attempt to rebuild the failed disk.

 

If nothing on the server is important or irreplaceable then I guess we will just go ahead and see what happens.

 

Currently, you have no parity protection since you already have a failed disk. Rebuilding to a new disk would at least get you back to a place where you had some redundancy so I would start with that.

9 minutes ago, trurl said:

In order to reliably rebuild every bit of a failed disk, Unraid must be able to reliably read every bit of parity PLUS every bit of ALL other disks. Parity by itself can recover nothing.

 

So obviously you don't want to wait until a disk completely fails to consider if any disks have problems since problems with any other disks that might not have completely failed yet can compromise any attempt to rebuild the failed disk.

 

If nothing on the server is important or irreplaceable then I guess we will just go ahead and see what happens.

 

Currently, you have no parity protection since you already have a failed disk. Rebuilding to a new disk would at least get you back to a place where you had some redundancy so I would start with that.

What about the unmountable drive though (disk 2)? Will unraid be able to use it to repair the new disk 3 drive?

My original question is about the order of operation since I have technically 2 drives that have failed, Disk 2 and Disk 3.

Edited February 27, 20215 yr by sloob

Whether you repair disk2 before or after rebuilding disk3 would have no effect on the rebuild. The SMART warnings you are already getting on disks 2 and 5 might be a problem though and nothing you can do about that except replace them one at a time after you get disk3 rebuilt. It is also possible that the SMART warnings you have on disk2 might cause problems repairing disk2. Just take things one at a time and see where we get.

 

Do you know how to replace a disk? Do you know what to expect during rebuild?

1 minute ago, trurl said:

Whether you repair disk2 before or after rebuilding disk3 would have no effect on the rebuild. The SMART warnings you are already getting on disks 2 and 5 might be a problem though and nothing you can do about that except replace them one at a time after you get disk3 rebuilt. It is also possible that the SMART warnings you have on disk2 might cause problems repairing disk2. Just take things one at a time and see where we get.

 

Do you know how to replace a disk? Do you know what to expect during rebuild?

I've only done it once and while I don't remember the procedure off the top of my head (I will be following the wiki as I do it) I remember it being relatively simple.

One thing I should mention is that when I originally had the problems of disks becoming corrupted/unmountable I was not able to access the shares via network. it only worked again when I repaired the drive with the -L option.

It's the case right now, I cannot connect the the SMB share.

Do you mean none of your shares are working, or just shares that are on disk2?

 

According to your diagnostics, user shares are mounted as are all disks except disk2, including the emulated disk3.