All,

 

I'm running unraid 6.5.3, and have 1 XFS disk 7(10tb) That is now unmountable with a "unmountable: no filesystem" message in the console.  This all occurred when there was a power outage and I rebooted the system and then started up unRaid.  Everything mounted fine, and the parity check begin to work. 24hours later I noticed the check was going extremely slow and this happened to me before due to lose wires.  So while the machine was up and running I was trying to push in the wires and the machine just powered off abruptly (I know I should have cancelled the parity and did a graceful shutdown before touch the machine's internals).  But when I brought the machine back up, I hadn't noticed a message saying there was an issue with my 1 XFS (can't be sure if it was an error on display or not)  but I just immediately started the array.  At that point I noticed  the XFS Disk 7 had the "unmountable: no filesystem" message.  I immediately stopped the array, and then ran an "xfs_repair /dev/sdi1" and that didn't fix it. So then I ran "xfs_repair /dev/sdi" and I forgot the "1".  So that ended up running for 24 hours and it couldn't find the secondary superblock.  I then again started the array hopin g the system would just emulate disk7 and I could see my data. But now it seems the parity is messed up and the emulation sees Disk 7 as "unmountable: no filesystem".   At that point I stopped the array and I realized I forgot the "1" and ran "xfs_repair /dev/sdi1" and it "fixed" the disk where at least I could mount it outside the array.  It does look like some of my data is missing(i think..its movies but can't be sure).  But now when I selected the now mountable drive for disk7, the array just thinks of it as a brand new disk.  Not sure what my next steps should be...

tower-diagnostics-20191212-1803.zip

1 hour ago, gustovier said:

I immediately stopped the array, and then ran an "xfs_repair /dev/sdi1" and that didn't fix it. So then I ran "xfs_repair /dev/sdi" and I forgot the "1".

Not the proper way to run file system repairs.  By accessing the drive directly, you've invalidated parity.  The proper way to run file system checks is via starting the array in maintenance mode and then

xfs_repair /dev/mdX

 

Thanks for pointing out the mistake. But you didn't offer what could be my next steps... 

Since parity is no longer accurate, I think your only recourse is to set a new config in the tools menu. Set the preserve all option, and correct any assignments before starting the array. The only disks that will be altered will be the disks assigned to the parity slots, they will be completely overwritten as parity is built again.

Since parity will be mostly in sync you can also check "parity is already valid" after the new config and before starting the array, then run a correcting check.

But how do I get the array to recognize my XFS drive 7 again?  It currently think it's a new disk when I assign the drive to disk7 and I do have data on it that I want to preserve

1 minute ago, gustovier said:

But how do I get the array to recognize my XFS drive 7 again?  It currently think it's a new disk when I assign the drive to disk7 and I do have data on it that I want to preserve

If you go the Tools->New Config route that has been suggested then when you start the array all the content of data disks previously used by Unraid is left intact.