BTRFS Uncorrectable errors help.

[rossiscool123]

Hello,

I've been having these random instances where my docker service just dies completely and the only way to recover is to hard reboot. My unRAID has been very unreliable, normally requiring a hard reboot every few days and I've had instances of all my containers just being removed. Its currently a setup of 2xNVME SSD's in RAID1 using BTRFS. I've copied all the files to another SSD and wiped the original drive and recovered from backup but still getting scrub errors.

 

I've read other posts regarding this topic saying to run  memtest to rule out if its faulty memory causing the corruption. Memtest ran for 3 hours and did two passes with no issue. I'm just asking here for advice on how to resolve these issues. I'll include my syslog and diagnosis output for someone to review and hopefully point me in the right direction.

cerberus-diagnostics-20240127-2349.zip syslog-192.168.1.200.log

[JorgeB]

Data corruption is being detected on both btrfs pools, this usually means a hardware issue, most often RAM related, if memtest didn't find any errors remove one stick of RAM, run a correcting scrub on both pools and if all errors are corrected reset the stats and keep monitoring the pools, if new corruptions are detected repeat the same with the other DIMM, this will basically rule out bad RAM.

[rossiscool123]

Backed up the drive to another drive and did the following https://docs.unraid.net/legacy/FAQ/check-disk-filesystems/#:~:text=So far there is not,a drive formatted with BTRFS).

Worked fine with no errors and all data is fine. Most of the BTFRS errors came from VM disks. Hope this helps anyone else with the same issue as me.