Since November I get an un-correctable error during fsck on a large ext4 file system placed on RAID5.
A 4 bay eSATA enclosure is hooked up to PogoPlug V4. Drives are WD Red 3TBx4.
Ext4 is set up with chunk size 256KB, stride 64 and stripe 192.
All drives pass extended SMART test , badblocks and WD read-write diagnostic test. RAID5 passes scrubbing with no problem.
Now, this very filesystem passes fsck under ArchLinux on PC without an issue!
BUMP. WOULD ANYBODY LOOK AT THIS? THIS IS A BUG BASED ON ALL I CAN SEE
However, under ARM Arch Linux I consistently get the error below. It seems that there is a kernel problem or a problem with mdadm package. Please let me know if I can help to collect logs to debug this.
Kernel : 4.9.0-1-ARCH
mdadm: 3.4-1
$this->bbcode_second_pass_code('', 'e2fsck -v -n /dev/md0
e2fsck 1.43.3 (04-Sep-2016)
/dev/md0 contains a file system with errors, check forced.
Pass 1: Checking inodes, blocks, and sizes
Error reading block 2147483680 (Attempt to read block from filesystem resulted in short read) while getting next inode from scan. Ignore error? no
Error while scanning inodes (268435456): Can't read next inode
')