I just noticed a few lines like this in my log:

 

Mar 31 09:05:42 Tower kernel: BTRFS warning (device loop0): csum failed ino 10426099 off 0 csum 1160006991 expected csum 1351314532

Mar 31 09:05:42 Tower kernel: BTRFS warning (device loop0): csum failed ino 10426099 off 0 csum 1160006991 expected csum 1351314532

Mar 31 09:05:42 Tower kernel: BTRFS warning (device sdh1): csum failed ino 114776 off 5693440 csum 2798177315 expected csum 3464834034

Mar 31 09:05:42 Tower kernel: BTRFS: read error corrected: ino 114776 off 5693440 (dev /dev/sdh1 sector 368572168)

 

I have two Samsung 850 Evo 500GB disks in a cache pool.  I looked at the attributes for each drive and neither are showing any reallocated sectors, hardware ECC recoveries, or any failures whatsoever.  Should I be concerned about this?  Is there something I can do to check the disks?

You can run a scrub on the cache pool webGui page and the docker page, docker image errors can't be fixed by scrubbing, if you start getting docker issues it's best to delete the image and create a new one.

 

 

 

 

No issues with my dockers... I ran a scrub on the cache pool and it came back with no errors.  So I have nothing to worry about then?

All should be ok, keep monitoring the log for a few days.

thanks!

Make sure you have a backup of everything on your cache.

 

I learned the hard way after csum errors popped up.

will do thx

Well, one of my VMs just went corrupt and while everything else seems to be running ok still on the cache pool my log file is scrolling like mad displaying the following:

 

What do I do?!?!

 

Mar 31 13:36:34 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:34 Tower kernel: BTRFS: sdh1 checksum verify failed on 37864603648 wanted 289BEC7B found 5709E6EC level 0

Mar 31 13:36:34 Tower kernel: BTRFS: read error corrected: ino 1 off 37864603648 (dev /dev/sdh1 sector 6714368)

Mar 31 13:36:34 Tower kernel: BTRFS: read error corrected: ino 1 off 37864607744 (dev /dev/sdh1 sector 6714376)

Mar 31 13:36:34 Tower kernel: BTRFS: read error corrected: ino 1 off 37864611840 (dev /dev/sdh1 sector 6714384)

Mar 31 13:36:34 Tower kernel: BTRFS: read error corrected: ino 1 off 37864615936 (dev /dev/sdh1 sector 6714392)

Mar 31 13:36:35 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:36 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:37 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:38 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:39 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:40 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:41 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:42 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:43 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:44 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:46 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

Mar 31 13:36:47 Tower kernel: BTRFS warning (device loop0): csum failed ino 28439 off 32768 csum 1919047842 expected csum 3148707344

 

Another thing I will add...  if I go to the main cache disk and do a scrub it runs through with no errors reported and nothing in the log.  If I go to the settings -> Docker page and run a scrub on the docker volume it still shows no errors in the unraid GUI however the log file starts scrolling the following (this is just a small bit of it):

 

I don't know what I should do.  Please help!

 

 

Mar 31 13:56:54 Tower kernel: BTRFS: checksum error at logical 4892672000 on dev /dev/loop0, sector 11669536, root 511, inode 24370, offset 192512, length 4096, links 1 (path: usr/share/libthai/thbrk.tri)

Mar 31 13:56:54 Tower kernel: BTRFS: checksum error at logical 4892672000 on dev /dev/loop0, sector 11669536, root 510, inode 24370, offset 192512, length 4096, links 1 (path: usr/share/libthai/thbrk.tri)

Mar 31 13:56:54 Tower kernel: BTRFS: checksum error at logical 4892672000 on dev /dev/loop0, sector 11669536, root 509, inode 24370, offset 192512, length 4096, links 1 (path: usr/share/libthai/thbrk.tri)

Mar 31 13:56:54 Tower kernel: BTRFS: checksum error at logical 4892672000 on dev /dev/loop0, sector 11669536, root 508, inode 24370, offset 192512, length 4096, links 1 (path: usr/share/libthai/thbrk.tri)

Mar 31 13:56:54 Tower kernel: BTRFS: bdev /dev/loop0 errs: wr 0, rd 0, flush 0, corrupt 31, gen 0

Mar 31 13:56:54 Tower kernel: BTRFS: checksum error at logical 4891406336 on dev /dev/loop0, sector 11667064, root 520, inode 24321, offset 184320, length 4096, links 1 (path: usr/share/icons/hicolor/icon-theme.cache)

Mar 31 13:56:54 Tower kernel: BTRFS: checksum error at logical 4891406336 on dev /dev/loop0, sector 11667064, root 519, inode 24321, offset 184320, length 4096, links 1 (path: usr/share/icons/hicolor/icon-theme.cache)

Mar 31 13:56:54 Tower kernel: BTRFS: checksum error at logical 4891406336 on dev /dev/loop0, sector 11667064, root 518, inode 24321, offset 184320, length 4096, links 1 (path: usr/share/icons/hicolor/icon-theme.cache)

Mar 31 13:56:54 Tower kernel: BTRFS: checksum error at logical 4891406336 on dev /dev/loop0, sector 11667064, root 517, inode 24321, offset 184320, length 4096, links 1 (path: usr/share/icons/hicolor/icon-theme.cache)

AFAIK docker image errors can't be fixed, best way is to delete the docker image, create a new one and re-add your dockers.

Sorry for the multiple posts but further troubleshooting has revealed the following:

 

I shut down all of my dockers and VMs.  I powered them on one by one and tested them.  The VM I thought was corrupt is actually fine.  Everything runs fine except my crashplan docker.  When I launch that one, the log file starts to scroll the csum failed warnings...  does that mean that this docker container is corrupt?  Shutting down crashplan and these warnings stop, so I'm starting to think it's just a corrupt docker container and the physical disks are ok?

 

I need crashplan though, so is the next logical thing to do to delete the crashplan docker and then re-install it?  If I leave the appdata folder in place, assuming I configure the crashplan docker the same after I re-install it, it should work as it was again?  Of is it the entire docker image that needs to be removed and start over again with all of my dockers?  The other dockers appear to be fine.

I'm not a docker expert, so wait for someone else to confirm, but I think you have to delete the docker image and create a new one, you should be able to add all your dockers from the user templates and keep all settings, the config files are all on appdata, not on the image itself.

Ok thank you... if anyone can confirm what I should do it would be much appreciated.

Since I am impatient (LOL) I went ahead and deleted the crashplan docker and image and then re-installed it.  After launching crashplan I am no longer getting csum failed warnings in the log.  Hopefully that did the trick but I'd still appreciate it if someone can confirm I'm good now or if I should still be deleting the entire docker image file and starting from scratch?

 

Thanks for everyone's help, I appreciate it.