So, I had a drive with a Red Meatball, shut down, checked cabling, rebooted, now I have a red meatball and a missing drive (missing drive is a 2tb wd green drive that has now proven bad, tons of errors on reboot, tested it, won't read), the red meatball drive tests fine with smartctl, but since the 2tb drive went missing (bad), I can't get the array online "too many missing/bad disks" error.  , I have a new drive in the system to replace the 2tb drive (a 3tb wd red) but since the first drive is still red meatball, it won't let me replace the 2tb missing with the 3tb, but I can replace the orignal red meatball with a new drive, and it turns blue, but still can't turn array on, due to missing 2tb drive...sorry if this is confusing.

 

 

Any suggestions?

Step one place all drives in correct slot..... Meaning make sure parity drive is correct in correct slot.

Step two start array with missing drive not assigned. Array should start and all data from drive should still show up... Please confirm very important.

Step 3 ( warning don't do this unless step two works data loss can happen)  stop array assign new 3tb drive as parity and old parity drive as the one that was missing.

Step 4 new copy button on bottom of page should appear. If so then you did steps correctly if not go to step one.

Step 5 press copy button and wait for copy to finish.

Step 6 start array this will start a rebuild of lost drive.

 

Hope this helps.

Thornwood

This is the correct way to do a parity swap.

I miss understood I thought you were trying to recover with a 3TB when your pairty was a 2TB sorry.

ok, not sure why I would want to replace the parity drive.

 

drive 5 shows red meatball (3tb drive) but tests fine, assuming write error or something threw it out of array.

drive 6 shows missing (2tb drive that is dead, really dead).

 

Can't start the array due to "too many missing/bad disks" so I have no way of getting it started to reassign the new 3tb drive to the 2tb spot, or even to the 3tb spot.  I have 2 new drives, so I could swap both out, but since I can't start the array, it won't let me unassign them and reassign them.

ok, not sure why I would want to replace the parity drive.

 

drive 5 shows red meatball (3tb drive) but tests fine, assuming write error or something threw it out of array.

drive 6 shows missing (2tb drive that is dead, really dead).

 

Can't start the array due to "too many missing/bad disks" so I have no way of getting it started to reassign the new 3tb drive to the 2tb spot, or even to the 3tb spot.  I have 2 new drives, so I could swap both out, but since I can't start the array, it won't let me unassign them and reassign them.

 

I think Thornwood missed the fact that you are out 2 drives not just one.

here what I would suggest but thread carefully here.

it is probably obvious that you lost one drive worth of data already.

as your 2TB is dead.

first and foremost take the red-balled 3TB out and using other pc you can read it on , copy ALL the data on it somewhere safe.

this way you would only loose one drive worth of data(the 2TB) at most.

try moving the disk to another SATA port and use new cable for it. if you are running v5.0 it should work ok.

if it's the cable or the port or both you will see it right away.

red ball means hardware problem and if you sure the drive ok than it's cable or sata port.

if it works, than you should be able to rebuild your bad 2TB drive.

otherwise you will still loose the data on that drive but on this one drive only

sorry...

 

 

 

I don't have any experience with doing this, so I'd try to solicit advice from someone who has, but it's my understanding that you should be able to set a new configuration with all your drives except the truly dead one, and somewhere there should be a setting that tells unraid to trust that your parity is still ok. That way you may be able to rebuild the dead drive using parity and all the other drives. Problem is, the red balled drive is guaranteed to have wrong data on it, so the rebuilt disk may not be readable without checking the file system or other magic. I wouldn't give up on the data just yet.

I don't have any experience with doing this, so I'd try to solicit advice from someone who has, but it's my understanding that you should be able to set a new configuration with all your drives except the truly dead one, and somewhere there should be a setting that tells unraid to trust that your parity is still ok. That way you may be able to rebuild the dead drive using parity and all the other drives. Problem is, the red balled drive is guaranteed to have wrong data on it, so the rebuilt disk may not be readable without checking the file system or other magic. I wouldn't give up on the data just yet.

 

This is correct. But first you may want to put disk 5 in a windows machine and use a recovery utility to copy it's contents. You can try this with disk 6 as well.

 

Set a new config that includes a the drives that are responding including a new drive in slot 6. Check the "parity is correct" box and start the array. Now stop the array and un-assign disk 6 then start the array and stop it again. Now assign disk 6 and its contents should be rebuilt.

 

 

I don't have any experience with doing this, so I'd try to solicit advice from someone who has, but it's my understanding that you should be able to set a new configuration with all your drives except the truly dead one, and somewhere there should be a setting that tells unraid to trust that your parity is still ok. That way you may be able to rebuild the dead drive using parity and all the other drives. Problem is, the red balled drive is guaranteed to have wrong data on it, so the rebuilt disk may not be readable without checking the file system or other magic. I wouldn't give up on the data just yet.

 

This is correct. But first you may want to put disk 5 in a windows machine and use a recovery utility to copy it's contents. You can try this with disk 6 as well.

 

Set a new config that includes a the drives that are responding including a new drive in slot 6. Check the "parity is correct" box and start the array. Now stop the array and un-assign disk 6 then start the array and stop it again. Now assign disk 6 and its contents should be rebuilt.

 

but his disk 5 is red-balling  and disk 6 is dead

I tought UnRaid would not allow for 2 disk failure.

are you sugesting that he still might be able to recover the dead disk data if he drops disk5 compleatly and redo array with new disk inplace of 6?

 

 

I don't have any experience with doing this, so I'd try to solicit advice from someone who has, but it's my understanding that you should be able to set a new configuration with all your drives except the truly dead one, and somewhere there should be a setting that tells unraid to trust that your parity is still ok. That way you may be able to rebuild the dead drive using parity and all the other drives. Problem is, the red balled drive is guaranteed to have wrong data on it, so the rebuilt disk may not be readable without checking the file system or other magic. I wouldn't give up on the data just yet.

 

This is correct. But first you may want to put disk 5 in a windows machine and use a recovery utility to copy it's contents. You can try this with disk 6 as well.

 

Set a new config that includes a the drives that are responding including a new drive in slot 6. Check the "parity is correct" box and start the array. Now stop the array and un-assign disk 6 then start the array and stop it again. Now assign disk 6 and its contents should be rebuilt.

 

but his disk 5 is red-balling  and disk 6 is dead

I tought UnRaid would not allow for 2 disk failure.

are you sugesting that he still might be able to recover the dead disk data if he drops disk5 compleatly and redo array with new disk inplace of 6?

 

 

 

No. Disk 5 will no longer be red balled once a new config is set. Only with disk 5 and all on the other drives in place can an attempt be made to rebuild disk 6.

but his disk 5 is red-balling  and disk 6 is dead

I tought UnRaid would not allow for 2 disk failure.

are you sugesting that he still might be able to recover the dead disk data if he drops disk5 compleatly and redo array with new disk inplace of 6?

No, a new config with a trust parity option will force unraid to use the contents of the currently red balled disk 5, along with all the other drives including parity, to possibly reconstruct at least some of the completely dead disk 6. The red ball on 5 just means that at some point, unraid couldn't successfully write to it, so it wrote the difference to the parity disk instead, and disabled disk 5. Whatever write failed, and all subsequent writes to the emulated disk 5 will cause the reconstruction of disk 6 to be wrong, but only at those addresses. Taking into account the robustness of reiserfs, it's entirely possible that the contents of disk 6 may be able to be recovered.