So I woke up this morning to a couple of emails from my monsters server telling me a disk had errors, when I check the server it indeed has a red x and a parity rebuild is in process. I am running 6.2b23 with two parity disks, so I'm guessing I'm ok, will have to get a replacement drive though. Fortunately there was nothing on the drive that has the errors. I have attached diags if anyone wants to have a look. I guess its part of the beta that the parity rebuild started automatically??

 

https://www.dropbox.com/s/hd0vvsv1v06sz8q/tower-diagnostics-20160621-0928.zip?dl=0

 

I guess its part of the beta that the parity rebuild started automatically??

 

Nope, a parity check started at midnight and disk25 dropped offline 5 minutes later, that check is still running.

So should I let the parity check finish? As I said there was nothing on the drive that dropped offline.

I thought since the drive dropped off, the parity check itself would be invalid.

I thought since the drive dropped off, the parity check itself would be invalid.

With dual parity it should be possible to check parity with only one disabled disk.

Was that a scheduled parity check? Correcting?

 

Don't know why your diagnostics don't have  SMART report files for all of your drives, maybe something flaky about the beta. There are only 14 files in the smart folder. But the one johnnie said, disk25 has a SMART report file but no actual SMART report in it. So we don't know if there is anything wrong with the drive or not.

 

I think I would cancel the parity check, shutdown, and check connections to see if you can get another smart for that drive.

Was that a scheduled parity check? Correcting?

 

Don't know why your diagnostics don't have  SMART report files for all of your drives, maybe something flaky about the beta. There are only 14 files in the smart folder. But the one johnnie said, disk25 has a SMART report file but no actual SMART report in it. So we don't know if there is anything wrong with the drive or not.

 

I think I would cancel the parity check, shutdown, and check connections to see if you can get another smart for that drive.

 

He's using and Areca that may not pass the SMART info, dropped disk is probably connect to it.

The parity check was indeed scheduled. The disk that has dropped off is not connected to the Areca. So you guys figure I should cancel the  parity check, shut down, check connections (which I know to be fine) reboot and run a SMART test on the drive and upload the diags?

Probably best 

If I wanted to just remove the drive from the array and not replace it, so reduce the size of the array from 100TB to 95TB, thats possible, correct? I'd just have to run a parity rebuild right?

If I wanted to just remove the drive from the array and not replace it, so reduce the size of the array from 100TB to 95TB, thats possible, correct? I'd just have to run a parity rebuild right?

Yes. New Config and then let it rebuild parities. Be sure you assign the drives correctly. If you accidentally assign a data drive to a parity slot it will get overwritten with parity.

I'm not sure whats going on, I powered down, checked the connections, pulled the drive, connected it to another Unraid machine that saw it just fine, put it back, booted up, it sees the drive, SMART test passes, it just doesn't like it for some odd reason.

 

Diags attached.

 

tower-diagnostics-20160621-1217.zip

... it just doesn't like it for some odd reason...

What do you mean by this? Are you rebuilding the drive?

No I'm not doing anything, the server is up, the array is mounted, the drive has a red x beside it but it sees the drive, I can get into the SMART data, there are no errors, the array is not rebuilding presently. Not sure what to do next.

After a disk is disable it has to be rebuilt, to a new disk or the old one, if you think the old is fine.

So how do I rebuild the disk? Do parity check or?

What do I do if I get a red X next to a hard disk?

Did that and now its rebuilding parity.

 

Thanks, will let you know how it goes.

 

Did that and now its rebuilding parity.

 

Thanks, will let you know how it goes.

I hope you mean it is rebuilding data, not parity.

Yes rebuilding data, although there was nothing on the drive to rebuild.

Yes rebuilding data, although there was nothing on the drive to rebuild.

In that case I guess you could have removed it then rebuilt parity instead. But assuming you want to keep the drive in the array it needs to be rebuilt. Did it ever have anything on it, or did you try to put something on it and then it failed? unRAID only disables a disk when a write to it fails. If anything was written to the disk while it was disabled then those writes will be rebuilt from parity.

No the drive has been in my array for well over a month, maybe longer, it along with a few other drives were added while I still had plenty of space elsewhere on the array for UnRaid to put data so it just simply never got around to them before it mysteriously dropped off.

I think what might have happened is unRAID tried several times to read the disk while it was checking parity, then when it couldn't read the data from the disk it attempted to write the missing data back to the disk from the parity calculation, then when the write also failed the disk was disabled.

I think what might have happened is unRAID tried several times to read the disk while it was checking parity, then when it couldn't read the data from the disk it attempted to write the missing data back to the disk from the parity calculation, then when the write also failed the disk was disabled.

Correct, whenever unRAID can't read data from a disk it uses parity and all other disks to try and write the data back, if it succeeds the read errors appear on the main page on the errors column but the disk remains enable, if the write fails the disk is disabled.

 

Parity rebuild was successful all is good now, thanks for your help.