Yesterday, I noticed that 3 of my 4 data drives were overheating due to constant a constant read from a program running across the network. I stopped the program causing the reads. Shortly after I did the unraid became unresponsive, the WebGUI wouldn't load and I could ping the tower but that's it. I was offsite so I couldn't physically check the console or restart the machine. 

 

This morning I come in and forced it to restart, now it's not mounting any of the drives except for what seems to be the parity. Starting the array causes the WebGUI to freeze, and I attached a picture of the what the console says shortly after. 

 

I need this resolved hopefully today, as this contains all of my company's important documents. 

 

diagnostics-20170321-0843.zip

Edited March 21, 20179 yr by techkitten

Check filesystem on disk1 (md1)

 

https://lime-technology.com/wiki/index.php/Check_Disk_Filesystems#Drives_formatted_with_XFS

3 minutes ago, techkitten said:

Yesterday, I noticed that 3 of my 4 data drives were overheating due to constant a constant read from a program running across the network

 

This would imply poor cooling, no disk went above 53C, so a little on the high side but far from critical.

8 minutes ago, johnnie.black said:

Check filesystem on disk1 (md1)

 

https://lime-technology.com/wiki/index.php/Check_Disk_Filesystems#Drives_formatted_with_XFS

I started a check on disk1, I seem to have had to check the systems a lot on all of my drives lately. I think this is the 2nd time in the last month, would there be something causing it to need checks more often?

Is the server having unclean shutdowns due to power cuts or crashing? That's the number one reason for filesystem corruption, if it's not there could be a hardware issue, like bad RAM.

No unclean shutdowns that I know of, if it has to be shutdown I normally initiate it through the webgui. But I will check the ram as I didn't know that it could cause filesystem corruption. 

End result of the xfs_repair.

I stopped the array, rebooted the tower, and successfully started the array! 

Thank you for helping me to resolve the problem!

30 minutes ago, techkitten said:

But I will check the ram as I didn't know that it could cause filesystem corruption. 

Obviously bad RAM can corrupt anything and everything, since all data goes through it.

1 hour ago, techkitten said:

This morning I come in and forced it to restart,

 

Also wasn't this an unclean shutdown?

 

1 hour ago, johnnie.black said:

 

Also wasn't this an unclean shutdown?

 

It came backup like it was a clean one. I went through the console and used the poweroff command.

If a parity check didn't start it was a clean shutdown, in that case I would run a few Memtest passes.