Errors with my ssd cache drive

[lebnet]

Hi everybody,

 

I'm a unraid user since 3 years and this is the first time I post on the forum. but not the first time I read  thank you all 

 

This morning I had the following message when I was trying to find why my docker containers were inaccessible : 

"Docker Service failed to start."

 

so I started to investigate the logs and I found my cache drive was in "read-only mode".

 

I tried to reboot but without success. After two clean reboots, my ssd cache drive (btrfs) was unmountable "Unmountable: Wrong or no file system".

 

I started the array in maintenance mode to make some checks and I have the following logs :

 

[1/7] checking root items
[2/7] checking extents
data backref 23819890688 root 5 owner 214345345 offset 274432 num_refs 0 not found in extent tree
incorrect local backref count on 23819890688 root 5 owner 214345345 offset 274432 found 1 wanted 0 back 0x27039250
incorrect local backref count on 23819890688 root 5 owner 214345347 offset 274432 found 0 wanted 1 back 0x17abcf90
backref disk bytenr does not match extent record, bytenr=23819890688, ref bytenr=0
backpointer mismatch on [23819890688 4096]
data backref 28792696832 root 5 owner 112068549 offset 0 num_refs 0 not found in extent tree
incorrect local backref count on 28792696832 root 5 owner 112068549 offset 0 found 1 wanted 0 back 0x4341a480
incorrect local backref count on 28792696832 root 5 owner 112068551 offset 0 found 0 wanted 1 back 0x2d407300
backref disk bytenr does not match extent record, bytenr=28792696832, ref bytenr=0
backpointer mismatch on [28792696832 61440]
data backref 102934360064 root 5 owner 2160095 offset 200638464 num_refs 0 not found in extent tree
incorrect local backref count on 102934360064 root 5 owner 2160095 offset 200638464 found 1 wanted 0 back 0x42373520
incorrect local backref count on 102934360064 root 5 owner 2160095 offset 144115188276494336 found 0 wanted 1 back 0x42373650
backref disk bytenr does not match extent record, bytenr=102934360064, ref bytenr=0
backpointer mismatch on [102934360064 8192]
ERROR: errors found in extent allocation tree or chunk allocation
[3/7] checking free space tree
[4/7] checking fs roots
[5/7] checking only csums items (without verifying data)
[6/7] checking root refs
[7/7] checking quota groups skipped (not enabled on this FS)
Opening filesystem to check...
Checking filesystem on /dev/nvme0n1p1
UUID: aaaa-bbbb-ccccc-dddd //fake number
cache and super generation don't match, space cache will be invalidated
found 393069277184 bytes used, error(s) found
total csum bytes: 274458644
total tree bytes: 1600569344
total fs tree bytes: 908066816
total extent tree bytes: 339001344
btree space waste bytes: 365831926
file data blocks allocated: 1536289886208
 referenced 367089385472

 

Do you have some recommendations to help me make things better? 

tower-diagnostics-20220712-1342.zip

[JorgeB]

Since data corruption has been detected by btrfs fist thing I would recommend is running memtest, after that there are some recovery options here.

[lebnet]

Thank you. At the moment, it doesn't smell well [emoji51]

[lebnet]

After 30h, I stopped the test (1500 errors). I repeated the tests but with each RAM individually and I was therefore able to eliminate the one that was faulty.

 

After reboot whithout the faulty one, the cache drive is still "Unmountable: Wrong or no file system". Is it best to do the recovery options you told me ?

 

Edited July 13, 2022 by lebnet

[JorgeB]

That's expected, damage is done, try the recovery options then reformat.