On 2025-12-12, Mik J <[email protected]> wrote: > > # disklabel sd3 > # /dev/rsd3c: > type: SCSI > disk: SCSI disk > label: Samsung SSD 980 > duid: 8cf4eb1722bfa480 > flags: > bytes/sector: 512 > sectors/track: 63 > tracks/cylinder: 255 > sectors/cylinder: 16065 > cylinders: 121601 > total sectors: 1953525168 > boundstart: 0 > boundend: 1953525168 > > 16 partitions: > # size offset fstype [fsize bsize cpg] > a: 1953525120 0 RAID > c: 1953525168 0 unused
offset for 'a' is wrong, should be at least 64 > I used chatgpt to help me to find what could I do to solve the "softraid0: > sd5 was not shutdown properly" recreate the partition and softraid properly, restore from backup, forget about chatbots for things like this, they have no understanding, just present plausible-sounding responses which might work or might totally screw you over. > and it told me this > dd if=/dev/zero of=/dev/rsd3a bs=1m count=4 you just wrote 4MB of zeroes over the start of your partition, where the softraid metadata (if i read the code correctly that starts 8K in and lasts for 64K), which includes the password-protected key used to encrypt the contents of the softraid. it was probably already slightly broken due to the bad offset but noe it's gone. > If I type > echo "mypassword" | /sbin/bioctl -s -c C -l /dev/sd3a softraid0 > it works > if I type > echo "mywrongpassword" | /sbin/bioctl -s -c C -l /dev/sd3a softraid0 > it says wrong password > > I would like to know if the situation is completely lost ? i don't know. perhaps there is some hope if the key is still in RAM and a softraid expert is reading. i would assume any non-backed-up data is gone. (in general it's a good thing to assume that any non-backed-up data may be lost at any time, and backups are only any good if you can restore from them).
