[6.9.0] (and previous releases)] a corrupted file on a btrfs filesyem doesn't return -EIO when user shares are used

I know this report is quite old but was this issue ever investigated or reproduced in 6.10+? 

It's still an issue, with v6.10 or even v6.11-rc, I assume that it happens because of FUSE and it might not be an easy fix from LT side.

23 hours ago, JorgeB said:

even v6.11-rc

Curiously with v6.11.0-rc3 there's no i/o error even when copying using a disk share, looks like a Samba issue, it stumbles but still copies a known corrupt file without error, it still works as it should if copying locally with a disk share.

 

root@Tower15:~# pv /mnt/disk2/test/1.iso > /mnt/disks/temp/1.iso
 395MiB 0:00:07 [75.3MiB/s] [========>                                   ]  9% ETA 0:01:04
pv: /mnt/disk2/test/1.iso: read failed: Input/output error
         

root@Tower15:~# pv /mnt/user/test/1.iso > /mnt/disks/temp/1.iso
3.93GiB 0:01:06 [60.5MiB/s] [===========================================>] 100%            

 

 

2 hours ago, JorgeB said:

Curiously with v6.11.0-rc3 there's no i/o error even when copying using a disk share,

I assume you meant a User Share in this case looking at the command output posted

8 minutes ago, itimpi said:

I assume you meant a User Share in this case looking at the command output posted

If you transfer using SMB with user or disk share there's no error, there is if you do it internally using a disk share, so disk share copying without error with SMB is new.

Just want to bump this issue as, based on the discussion here https://forums.unraid.net/topic/134119-btrfs-on-unraid-checksum-on-read-when-transfering-data-to-windows-pc/?do=findComment&comment=1220283,

it appears to still be relevant in the 6.12 beta.