Arch Linux

gentao

Member

Registered: 2020-12-27

Posts: 6

BTRFS bad data blocks after particion clone from bad disk to new

Hello.

One day I noticed that my old hard drive dying, and I decided to use new hard drive. I copied  my ext4 partition with Arch Linux from old disk to new. ...unfortunately, not without errors. But Arch Linux works fine, and during month  I didn't noticed corrupted data blocks related problems.

Now I moved Arch Linux from ext4 to BTRFS. Moved without errors, works fine. ...until I didn't executed # btrfs balance start /
I got

ERROR: error during balancing '/': Input/output error
There may be more info in syslog - try dmesg | tail

# dmesg says

[ 1490.222404] BTRFS info (device sdb3): balance: start -d -m -s
[ 1490.224291] BTRFS info (device sdb3): relocating block group 121626427392 flags metadata
[ 1496.998086] BTRFS info (device sdb3): found 9811 extents, stage: move data extents
[ 1497.850578] BTRFS info (device sdb3): relocating block group 120552685568 flags metadata
[ 1511.520331] BTRFS info (device sdb3): found 65356 extents, stage: move data extents
[ 1512.616704] BTRFS info (device sdb3): relocating block group 118237429760 flags data
[ 1512.770123] BTRFS warning (device sdb3): csum failed root -9 ino 407 off 0 csum 0xd5ffab47 expected csum 0xdafecdf3 mirror 1
[ 1512.770132] BTRFS error (device sdb3): bdev /dev/sdb3 errs: wr 0, rd 0, flush 0, corrupt 3, gen 0
[ 1512.770167] BTRFS warning (device sdb3): csum failed root -9 ino 407 off 1048576 csum 0x9be2f7c3 expected csum 0x92599534 mirror 1
[ 1512.770175] BTRFS error (device sdb3): bdev /dev/sdb3 errs: wr 0, rd 0, flush 0, corrupt 4, gen 0
[ 1512.781896] BTRFS warning (device sdb3): csum failed root -9 ino 407 off 0 csum 0xd5ffab47 expected csum 0xdafecdf3 mirror 1
[ 1512.781905] BTRFS error (device sdb3): bdev /dev/sdb3 errs: wr 0, rd 0, flush 0, corrupt 5, gen 0
[ 1512.849819] BTRFS info (device sdb3): balance: ended with status: -5

# btrfs scrub start / says

UUID:             850545f4-8b58-42ee-97e4-f49b4fbfe550
Scrub started:    Tue Mar 16 21:55:05 2021
Status:           finished
Duration:         0:11:50
Total to scrub:   65.63GiB
Rate:             94.64MiB/s
Error summary:    csum=259
  Corrected:      0
  Uncorrectable:  259
  Unverified:     0

I checked RAM for errors by executing # memtester 200M 1, but all is ok.

The question is: Can I recover these bad blocks in new hard drive? If yes, then how?
If you need more information about my situation, please, ask.
Thank you for time spent to read my theme. I will be glad to any help.

Last edited by gentao (2021-03-17 12:10:53)