I've started to receive errors that one of my cache drives (btrfs cache pool) has pending sectors.

 

What is the best course of action to take?

Can the cache pool be reduced to the single drive that does not have any errors?

 

SMART report and diagnostics attached.

 

Many thanks.

tower-smart-20161223-1112.zip

See FAQ for unRAID V6 stickied at the top of this subforum. Once you have the problem drive removed from the pool you can preclear it to try and get the pending sectors reallocated.

I'm happy to remove the faulty drive from the pool and preclear, but I don't want to proceed any further without ensuring that the data on the drive pool is still usable. Should the good drive mount correctly when btfrs formatted as a single drive?

 

I have already tried stopping the array and allocating only the good cache drive, but it fails to mount.

 

Thanks.

So did you follow the instructions in the FAQ for removing a drive from the pool?

Hangs head in shame.

 

No, I'll do it when I get home and report back. Thanks for your advice.

Finally found time to get back to this. My current cache drive is a pair of 500Gb HGST drives (sdl and sdm) and one of the drives has pending sectors (48).

 

Following the guide, I have backed up the current cache and now have a pre cleared 500Gb replacement disk in the machine, ready to replace the failing disk.

 

I have issued the btrfs filesystem df /mnt/cache command and it looks like there is more than one data profile, so I am now looking for further help:

 

root@Tower:~# btrfs filesystem df /mnt/cache
Data, RAID1: total=136.00GiB, used=116.56GiB
Data, single: total=1.00GiB, used=0.00B
System, RAID1: total=32.00MiB, used=48.00KiB
Metadata, RAID1: total=3.00GiB, used=682.70MiB
GlobalReserve, single: total=240.00MiB, used=0.00B

 

Thanks.

tower-smart-20170203-1438.zip

A balance should fix that but since one disk has pending sectors and the single profile looks unused it's better to replace the problem disk first, used method 1 on the FAQ and the 4b option to replace a known bad disk, using the -r option to read data from the good disk only.

 

http://lime-technology.com/forum/index.php?topic=48508.msg484480#msg484480

 

 

Faulty cache drive successfully replaced following the guide. Many thanks for you help 

 

btrfs still shows the same 2 profiles as before. Should I do anything else to fix?

Run a balance using the GUI on the cache page using the default options.