lsblk & df giving different mount points

[AndyMH]

Title says it all.

lsblk:

Code: Select all

andy@T431 ~ $ lsblk -nio mountpoint /dev/sdb1
/run/timeshift/backup

df:

Code: Select all

andy@T431 ~ $ df | grep sdb1
/dev/sdb1                  1911193164  901146448 922710912  50% /media/ultrabay

sdb1 is the destination for my timeshift snapshots (it's my backup drive for timeshift, backintime and image backups). You can have the same partition mounted in different locations simultaneously, but why are the two commands different? The correct output is that both commands should show both mount points

[mikeflan]

Are they both listed in /proc/mounts ?

[AndyMH]

Yes

Code: Select all

andy@T431 /proc $ grep sdb1 mounts
/dev/sdb1 /media/ultrabay ext4 rw,relatime 0 0
/dev/sdb1 /run/timeshift/backup ext4 rw,relatime 0 0

[mikeflan]

I don't know. Just for grins lets see if this is different:

Code: Select all

df -h --output=source,target

I suspect it too will not list both mount points.

[AndyMH]

I suspect it too will not list both mount points.

Correct , just the /media/ultrabay mount point (which is mounted via fstab, not that I can think of any reason why this should make any difference). I assume timeshift is using mount, but why this should take precedence in the lsblk output, no idea?

[donten]

This tripped me up a few days ago...I opened Nemo, and the drive I use for backups was gone; timeshift had not released it. I took a quick look at the Disks utility, and the mount point was "run/timeshift" something. confused? That's not playing nice I had to boot from the Mint iso to be able to start timeshift and restore from one of the backups.

[donten]

Why does it say "confused," instead of "W T F?"

[Larry78723]

Could Timeshift be creating a softlink on the fly to the actual storage location (sdb1)?

[AndyMH]

Could be, but life's too short to start digging around in the source code . Pet peeve - virtually all the linux source code I've looked at has either zero or next to no comments. That's why foxclone has as many lines of comment as code.

[donten]

I checked again today, and timeshift did it again...But I got smart and simply unmounted the *hijacked* disk-- then I remounted it and Disks restored the original mount point of itself

[donten]

I found that by using the UUID of the drive as the mount point solves the issue.