So A few days back I noticed that one of my dockers was not started as it should be. The logs showed that the config directory wasn't where it should be, which I found odd because when I went to look for it, all the files showed up as they should. I thought this was an error with an update to the docker where maybe the actual application code was looking in the wrong spot, maybe a typo or something.

 

I left it as is and just didn't use the application for a day or two, until yesterday. I needed to reboot for something so I did, and while watching all the text fly-by I noticed that there was an error along the lines of BTRFS critical error: unable to find logical 123189231236 found 4096 (the numbers aren't exactly correct as they're off the top of my head) but I noticed anytime I enter the directory, the files are no longer there and that error pops up again.

 

First I thought that the SSD was failing, so, no biggie, I'd just install the spare I have which is also larger. In addition, all my searching online seems to back this up, and all the documentation I can find regarding that error says it's bitrot, or something. However, when I swapped the new drive in and copied my appdata over, I continue to get that error when I'm looking for the same file.

 

Lastly, I also goofed because I tried to open a 35gb zip file in krusader which seems to have massively inflated the size of my docker image so I thought I'd move it off the cache to the array. I followed steps to move it, and then decided to just rebuild the whole darn thing. Either way, I must have messed something up because now it shows I have 0 dockers installed, but really, I've got backups of the appdata, so that's no big deal, I'll just reinstall.

 

I just thought I would post here, and see if anyone has any suggestions because I can now no longer delete the directory that is "corrupt" as it shows an I/O error, so I'm wondering if anyone has experienced this, or knows a solution for how to even just delete the directory?

 

I'm at work, so I don't have access to my server to test but I thought I'd fish for suggestions here, and worst case, pull the cache, completely format it on my windows machine, and then start over, but bypass copying back the folder that I'm assuming is "corrupt",

Please post the diagnostics : Tools -> Diagnostics

7 minutes ago, johnnie.black said:

Please post the diagnostics : Tools -> Diagnostics

Thanks, will do as soon as i get home.

4 hours ago, johnnie.black said:

Please post the diagnostics : Tools -> Diagnostics

I've attached the diagnostics

thor-diagnostics-20190228-2321.zip

It's an unusual error, but there's metadata corruption so I would recommend backup, re-format and restore, you can use this if needed to help with the backup/restore.

 

Also a good idea to run memtest see if anything pops up.

Just now, johnnie.black said:

It's an unusual error, but there's metadata corruption so I would recommend backup, re-format and restore, you can use this if needed to help with the backup/restore.

 

Also a good idea to run memtest see if anything pops up.

Yea, I really only had appdata on the cache so I have that backed up. I even changed the drive over and then restored the data and it still has that error. I'm thinking my next step is to avoid the file that I suspect is causing the issue and then restore everything but that. My main concern is that hopefully my Plex, Sonarr, and Radarr data is fine, everything else is replaceable easily.

 

Thanks for the support!

Just in case someone else has this error, since I had a hard time finding any solution that wasn't just "Your SSD is dead" here is what I did to fix my situation: 

 

• As johnnie.black suggested, I ensured I still had my backup of my AppData (the only thing on my cache drive) and then I shut down the array.
• I removed the cache from the array (not physically, just in software)
• Using unassigned devices, I formatted the cache completely
• Re-added the cache back to the array and started it up again.
• In my situation, I had to rebuild my docker image by re-installing all the dockers individually.
• I then copied all the AppData back and overwrote all the AppData that was currently there from the fresh installs.

 

Again, this might be unique to my situation but I knew which file was corrupt because it actually caused a docker to fail, so when I copied all my files back, I left the errored file out (obviously enough, as the whole reason this was being done is to fix that corruption). It seems like it was a file that was corrupt that threw up an error that is the same as what might happen with a failing drive. I also swapped out my cache drive, just to be safe, but considering when I transferred the broken file it threw up the same error, I'm confident that it wasn't the drive with the issue, just that file.

 

Currently, the 40Gb of Plex metadata is still transferring back but everything else that was transferred seems to have worked flawlessly. I also spent a few hours running memtest, and after a pass, it came back as fine. I'm aware that you should really run it for more than just one pass, but I wanted to start the transfer. Once everything is confirmed to be back up and running, I plan on running memtest for the whole weekend and then seeing how it fairs.

 

Thanks for everyone here and on Reddit who helped with the solution!