Docker issues and error reports after recent MOBO firmware upgrade

[omartian]

Hi Guys-

 

Recently upgraded the firmware for my aorus b450 pro wifi to version f60e.  Had an issue where my unit wouldn't boot to bios.  It turned out that i could get into bios w/out my sas controller was unplugged.  I changed the pcie settings in the bios and got my server up and running again.  

 

Did a pre-clear of a 14 tb drive which took approximately 4 days and noticed i would get emails in the morning w/the following error code:

 

fstrim: /etc/libvirt: FITRIM ioctl failed: Input/output error
fstrim: /var/lib/docker: FITRIM ioctl failed: Input/output error

 

Would also get this code:

 

**** Unable to write to cache ****   **** Unable to write to Docker Image ****   **** unRaid's built in FTP server is running ****  

 

I've never precleared a disk before so i attributed it to the preclear process.  It took 4 days for the 14 tb drive but i got the ok from preclear that the disk was good.  I haven't received the "fstrim" error message since monday (did the preclear over the weekend), but still get the cache/dockerimage/ftp server messages the last few mornings.  

 

Also, another thing i've noticed is that my docker's (plex and binhex-krusader) run fine, but if i stop them, i can't restart them.  I get an "execution error, error code 403" message and both dockers refuse to start.  if i reboot my system or stop/start array, it'll come back up but i've never had this issue before. 

 

Attached is my diagnostics file. 

 

Do you think there can be an issue w/the sas controller for one or both of the error messages.  

 

Any and all help is much appreciated.  

nasgard-diagnostics-20210107-0840.zip

[Squid]

I think you need @JorgeB.

 

The way that I see it, btrfs usage shows ~931G of total space on the cache pool, but df is reporting that the size is only 466G, and this is playing havoc with the docker.img file.

[omartian]

That was for redundancy for the cache when i  first set it up a year ago, but i soon found a lot of my files getting stuck in the cache drives and would get constant notifications about how my cache drives were about to be full.  

 

Should i re-classify these drives?

[KZ7]

It looks like the filesystem on sdd1 is corrupt which ends up causing the docker images mounted on that filesystem to become readonly. The docker image errors are a result of the corruption on /dev/sdd1.
 

Jan  7 08:27:09 Nasgard kernel: BTRFS critical (device sdd1): corrupt leaf: root=2 block=948912128 slot=56, unexpected item end, have 134231043 expect 13315
Jan  7 08:27:09 Nasgard kernel: BTRFS: error (device sdd1) in __btrfs_free_extent:6805: errno=-5 IO failure

Jan  7 08:27:09 Nasgard kernel: BTRFS info (device sdd1): forced readonly

 

[JorgeB]

Yep, pool fs is corrupt, best bet is to backup and re-format.

[omartian]

ok cool.  i reformatted and re-added my cache drives.  hopefully it's not a failing disk and I don't see these error reports anymore.  thanks for your help.