[SOLVED] migrate Data from Failed drive and remove from raid?

[oryhara]

Dear UnRAID community,

I have a failed drive, and more than plenty of space on the remaining drives to migrate it off. However, I have several shares which span all of the drives in the RAID and do not know the best way to proceed in such a way that will best preserve my data. 

 

Parity is valid, and the data on the failed drive is being reproduced from the parity. 

 

All the drives in the RAID are 1TB(various manufacturers and various ages).  I would just buy a replacement for the failed drive, but the prices on drives are more than twice what they were 2 months ago, so I'm not inclined to do that.  I have a 2TB in an external enclosure that I've been using to travel, but it is larger than the parity drive, so i cannot insert it into the raid, and given the failed state of the other drive, i cannot replace the parity drive. 

 

My question is: how can i move the data from the failed drive to other 1TB drives(none empty, but at several TB of free space between them), then remove the failed drive from the RAID and restore the parity?

 

Ive searched the forums and the wiki, and found several articles and topics similar to this, but none that I felt accurately covered the whole picture in a way that gave me enough confidence to proceed, at least not when my hoard is at risk. 

[Joe L.]

Dear UnRAID community,

I have a failed drive, and more than plenty of space on the remaining drives to migrate it off. However, I have several shares which span all of the drives in the RAID and do not know the best way to proceed in such a way that will best preserve my data. 

 

Parity is valid, and the data on the failed drive is being reproduced from the parity. 

 

All the drives in the RAID are 1TB(various manufacturers and various ages).  I would just buy a replacement for the failed drive, but the prices on drives are more than twice what they were 2 months ago, so I'm not inclined to do that.  I have a 2TB in an external enclosure that I've been using to travel, but it is larger than the parity drive, so i cannot insert it into the raid, and given the failed state of the other drive, i cannot replace the parity drive. 

 

My question is: how can i move the data from the failed drive to other 1TB drives(none empty, but at several TB of free space between them), then remove the failed drive from the RAID and restore the parity?

 

Ive searched the forums and the wiki, and found several articles and topics similar to this, but none that I felt accurately covered the whole picture in a way that gave me enough confidence to proceed, at least not when my hoard is at risk. 

easiest is to enable "disk" shares and then copy the files from the simulated failed drive to other "disk" shares in the array.

To minimize confusion, you can temporarily disable the user-shares.  You can use window's explorer to copy the files, or log onto unRAID and use "mc"

 

Then, once you have copied all the files off of the (simulated) failed drive, you need to stop the array, un-assign it, set a new disk configuration, (which immediately invalidates parity) and then start the array with the disk remaining un-assigned.  unRAID will then calculate party on the new disk configuration.

[oryhara]

ok, so i copied the data off of the bad drive and reassigned my drives after an initconfig.

Parity calc is all set again but now when i load the shares, the data which was on the failed drive does not appear.  If i telnet into the unraid box then i can plainly see the data, but the share appears to still be loading the files as if they were on the drive which had failed(and no longer is in the RAID).

 

[dgaschk]

Are you running a 5.0 beta? if so, run the New Permissions script.

[oryhara]

beginning script run now. 

 

i did a little looking.  the data is clearly visible when i browse the files on the unRAID web GUI, but in both SMB and AFS a number of user share directories appear to be empty.  not the root of the user share, but at least 3 layers deep within it.