Disk Read Issues

[0xSeb]

Hello everyone! 

 

I have recently switched from TrueNAS over to Unraid and have been loving the experience! 
I didn't bring over any data and started things from scratch as I have been trying to find a platform that suites my needs best. 

 

I have been running this server for about a week, and it seems that Disk 1 has thrown a few read errors which caused the array to stop. 

 

I ran an extended SMART test with zero errors as well as a filesystem check on the drive (which seems to have zero errors, but I could be wrong...). 

Before these drives were put into service (before the transition to Unraid), I ran a badblocks session as well as an extended SMART test before and after with no errors. 

 

Attached is the diagnostics as well as the SMART test on the one drive, but let me know if more info is needed. 

I am currently running a read check on the array in maintenance mode to check the files on the drives. 

 

Any advice on how to move forward would be greatly appreciated! 
 

ST8000VN004-3CP101_WWZ14C12-20230509-1145 disk1 (sdc) - DISK_DSBL.txt tower-diagnostics-20230509-1145.zip

[JorgeB]

Diags are after rebooting so we can't see what happened, also you are virtualizing Unraid, that is not supported, but for best results at least pass-through a HBA/controller.

 

 

[0xSeb]

New Diags below immediately after a reboot. 

 

Completely understood in regards to virtualizing. I have manually passed through all SATA ports directly to the VM running Unraid. As I'm in this transition, I'm noticing that Unraid can handle everything I need much easier than my prior setup, and am really considering just moving this to bare metal, but let me know your thoughts! 

 

Thanks! 

tower-diagnostics-20230509-1209.zip

[JorgeB]

32 minutes ago, 0xSeb said:

New Diags below immediately after a reboot. 

I think you misunderstood, the 1st diags were already after a reboot, and the syslog starts over after every boot, so cannot see the errors.

 

 

 

 

 

 

[0xSeb]

Sorry about the misunderstanding and I see what you mean now. 
 

Is there any advice on how to proceed from here?  
 

I don’t have any unrecoverable data that is super important to me aside from Docker containers and their respective configs at this point in time. 

[0xSeb]

As I have been doing more reading, I ran a filesystem check on the drive, and there was a small issue it seemed. Following documentation on steps to fix this on the Unraid wiki, things were repaired in a couple of minutes. 

I then went to remount the drive to the array, and in order to bring it back to an active state, I would need to do a data rebuild, but as there were no apparent issues with Disk 2 in my setup or my parity drive, it seemed like the next way to go. 

 

I ran the data rebuild, and after a handful of hours it started to throw read errors on disk 2 as well as the parity drive. I am unsure of what has occurred to cause this, but my assumption is this is due to whatever issue is caused by Disk 1's errors? 

 

I have attached Diags from this occurrence. 

tower-diagnostics-20230510-0005.zip

[JorgeB]

Like mentioned if you want to keep virtualizing Unraid you should pass-through a controller, running it like you are with virtual controllers is prone to cause disks errors like the ones you are seeing.

[0xSeb]

Understood! 

 

Data seems to be intact. I am currently moving it to the same hardware, but not on a virtual machine. I have tried just booting from the USB on bare metal (same machine) and all seems to be intact aside from asking for a parity sync. 

Should I start from absolute scratch here, or would a parity sync be a better place to begin to see if any errors appear? 

[JorgeB]

Parity sync should be fine.