drive failure notification?

[jtack]

I noticed my unraid box was a bit louder than usual even though there wasn't much happening.  Checked the unraid log and a billion messages saying:

 

Dec  9 22:32:02 marvin kernel: critical medium error, dev nvme0n1, sector 78259464 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 0
Dec  9 22:32:03 marvin kernel: nvme0n1: I/O Cmd(0x2) @ LBA 78259304, 8 blocks, I/O Error (sct 0x2 / sc 0x81) MORE DNR

At the same time:

Dashboard > Array > SMART says healthy.

Main > Pool Devices > cache drive nvme0n1 says 0 errors.

No warning/error pop-ups in the GUI.

 

I'm familiar with SMART and know not to rely it it to tell me where there are problems.  Is the Error count on the Main page only showing SMART errors?

 

How can I make sure I get notified about future problems like this?

 

 

 

 

 

 

 

[JorgeB]

If the device is in a btrfs pool you can use this:

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=700582

 

[jtack]

Thank you for that tip!  I've added the script for next time.

 

Now I'm trying to figure out how to fix this without making things worse.  Unraid is still pretty new to me...

 

I currently have two Pool Devices:

 

Cache_nvme (the broken one)

-btrfs

-1tb

-currently reporting 500+ read errors.  Zero write/flush/corruption/generation errors.  Since it's nvme and not sata there is no cable to fail/replace.  But I'll probably still try to re-seat it before buying a replacement.

 

Cache_ssd

-xfs

-512gb

-Zero errors

 

From what I gather by reading the FAQs I should be able to replace the failing btrfs drive and I'll only lose the appdata which only lives on that drive.  I still don't understand why the appdata folders on both pool drives aren't the same.

 

Once I replace the failed drive, does it make sense to reformat the xfs cache and change it to btrfs?

[JorgeB]

18 hours ago, jtack said:

I still don't understand why the appdata folders on both pool drives aren't the same.

They would only be the same if they were a mirror in the same pool

 

18 hours ago, jtack said:

Once I replace the failed drive, does it make sense to reformat the xfs cache and change it to btrfs?

As you prefer, single cache can be xfs or btrfs, depending on the user preference, both have advantages and disadvantages.