After pre-clearing Parity 2 I suddenly see this in the Main screen.  The drives (1 and 4) show Unmountable so I went to see if they could be repaired.  XFS gives the following error:

 

 

I have backups of everything but I'm wondering if my array is toast?  Or is there something I can do to get the drives working again?  SMART says they are ok but I am currently running an extended test.

The suspicion would be that the drive dropped offline which is why no SMART report is available.  If so then a reboot should get it back.  If possible get a diagnostics (Tools->Diagnostics) before rebooting so that you can post it here for analysis.

 

Assuming it comes back the question becomes why it dropped offline.  Commonest culprits are loose cables and/or inadequate power. 

I unfortunately didn't get a diagnostic on it before I rebooted.  After multiple reboots it's still the same (unmountable) and xfs_repair always failes with the I/O error, so I'm running an extended SMART test on it right now.

Where exactly does it say Unmountable? The red X just means disabled. And since you don't yet have dual parity I don't think the drives can be emulated, so maybe you can't actually repair the filesystem on the emulated md device and the message you got isn't really relevant or correct.

Where exactly does it say Unmountable? The red X just means disabled. And since you don't yet have dual parity I don't think the drives can be emulated, so maybe you can't actually repair the filesystem on the emulated md device and the message you got isn't really relevant or correct.

 

You can't run xfs_repair on the emulated disks because you have 2 and only one working parity.

 

Assuming SMART is OK for both disks, you should do a new config with single parity only using the trust parity option, then do a parity check and if everything is OK add parity2.

OK.

 

I still have to wonder about this situation though. You don't have valid parity2. You have 2 disks disabled (or at least with red X which is what it is supposed to mean).

 

Is it actually trying to emulate these disks even though you don't have both parity valid? And if it isn't trying to emulate the drives, then what is it trying to repair in the webUI?

 

The emulated drives would be md# and I thought that is what it would be trying to repair in the webUI. The actual disk partitions I think have to be repaired at the command line by referring to them as sdX1.

 

Could be the actual disk partitions aren't really unmountable, but the emulated disks are unmountable because parity2 isn't valid.

 

This may be a question for Limetech. Or at least a question for johnnie.black. Who I see already answered all this while I was typing!

You can't run xfs_repair on the emulated disks because you have 2 and only one working parity.

 

Assuming SMART is OK for both disks, you should do a new config with single parity only using the trust parity option, then do a parity check and if everything is OK add parity2.

And check your connections!

Assuming SMART is OK for both disks, you should do a new config with single parity only using the trust parity option, then do a parity check and if everything is OK add parity2.

 

Would that recover the data or is that toast? (Like I said I have backups)

 

Also, when you say "new config with single parity using the trust parity option" I'm not sure how to do that.  Is there a tutorial for what you want me to do?  Thanks for all the help.

New config should keep all your data:

 

-tools -> new config -> select retain disk = all and click aply

-back on the main page unassign parity2

-before starting the array check "parity is already valid" nest to the start button

-start array and do a parity check

 

Then reassign parity2

Could be the actual disk partitions aren't really unmountable, but the emulated disks are unmountable because parity2 isn't valid.

 

This may be a question for Limetech. Or at least a question for johnnie.black. Who I see already answered all this while I was typing!

 

Good question, in theory one of them should be correctly emulated, never tested this situation but will when I can.

 

Doh! Wrong, it can't emulate because there are two missing disks and only one valid parity, I was right the first time.

Seems like the webUI shouldn't even let you attempt a filesystem repair in this situation.

Maybe would should go ahead and get a diagnostic anyway. Perhaps one or both of those disks have issues and unRAID tried to rewrite (from parity calculation) something that was unreadable and that failed resulting in disabled disk(s).

I believe the OP rebooted and didn't save the diags.

I know, and he did say SMART was OK so maybe nothing to worry about.

I know, and he did say SMART was OK so maybe nothing to worry about.

 

I rebooted before pulling the diagnostics so I don't have them and I apologize.  As for the SMART tests, I am currently running Extended tests on both drives.  Short tests passed with no errors right before I started the long test.  I'm hopeful I can recover the array and I am very appreciative of all the hep that's been provided.

 

PS:  When I rebooted, it started a parity check\sync.  Would that effect the parity check when I create a new config with the drives now accessible? What I'm asking is that when I rebooted was it recalculating parity or just syncing Parity 1 and Parity 2?

I know, and he did say SMART was OK so maybe nothing to worry about.

 

I rebooted before pulling the diagnostics so I don't have them and I apologize.  As for the SMART tests, I am currently running Extended tests on both drives.  Short tests passed with no errors right before I started the long test.  I'm hopeful I can recover the array and I am very appreciative of all the hep that's been provided.

 

PS:  When I rebooted, it started a parity check\sync.  Would that effect the parity check when I create a new config with the drives now accessible? What I'm asking is that when I rebooted was it recalculating parity or just syncing Parity 1 and Parity 2?

Interesting question that perhaps johnnie.black has some insight to give us.

 

If it was a clean shutdown, and parity was valid, but parity2 wasn't synced yet, then it might have just been syncing parity2. I don't think the webUI gives clear indications of the different possible scenarios with 2 parity disks.

It could only be syncing parity2, but probably a bug since it can't be correctly synced with 2 disable disks.

Great news!  After the extended SMART tests finished with no errors\issues, I rebuilt the config per johnny.black and everything came up just like it was.  I removed Parity 2 and will start a Parity Check shortly.  I'll report back once that's started with a status.  Again, thank you to all that helped!