[SOLVED] 100% usage on /dev/mmcblk0p1

I added a 500GB SSD drive (formatted ext3) to the system (via the front USB port) and enabled it. The drive contains 7GB of content.

Unfortunately, this has resulted in the / directory being 100% full:

As a result, I can’t modify the .bashrc file nor start the minidlna service as a result.

Is this what is supposed to happen? What am I doing wrong?

Please have a look at my thread from yesterday: Disk space gone

Long story short: A snapshot is probably consuming all the disk space, use schapps list and schnapps delete ## to free some space.

Snapshots may play a role, but I think you just misplaced the data onto the internal memory instead of the SSD. If you look at the column „used“, that 7GB doesn’t show up on the SSD, it’s not there.

I would appear that you are correct, vorner.

Checked disk utilization:

root@turris:/# du -m / | sort -nr | head -n 20
7257    /
7139    /mnt/jetdrive
7139    /mnt
6253    /mnt/jetdrive/Movies
885    /mnt/jetdrive/Series/Humans
885    /mnt/jetdrive/Series
93    /usr
63    /usr/lib
14    /usr/bin
13    /usr/sbin
13    /usr/lib/opkg/info
13    /usr/lib/opkg
11    /usr/lib/python2.7
10    /lib
8    /lib/modules/4.4.13-05df79f63527051ea0071350f86faf76-9
8    /lib/modules
5    /etc
4    /sbin
3    /usr/share
3    /usr/lib/python2.7/lib-dynload

Physically removed the drive.

Checked /mnt/jetdrive:

root@turris:/mnt/jetdrive# ls
Movies  Music   Series

Deleted the directories:

root@turris:/mnt/jetdrive# rm * -rf
root@turris:/mnt/jetdrive# ls

Checked disk usage:

root@turris:/mnt/jetdrive# du -m / | sort -nr | head -n 20
118    /
93    /usr
63    /usr/lib
14    /usr/bin
13    /usr/sbin
13    /usr/lib/opkg/info
13    /usr/lib/opkg
11    /usr/lib/python2.7
10    /lib
8    /lib/modules/4.4.13-05df79f63527051ea0071350f86faf76-9
8    /lib/modules
5    /etc
4    /sbin
3    /usr/share
3    /usr/lib/python2.7/lib-dynload
2    /usr/lib/python2.7/site-packages
2    /usr/lib/python2.7/encodings
2    /usr/lib/lua/luci
2    /usr/lib/lua
2    /etc/ssl/certs

I manually created /mnt/jetdrive to be the mounting target. When I uploaded files, it must have been to that folder, rather than to the mounted drive.

What is the ‘right’ way to mount a read/write drive, such that it will remount when the system reboots and can be used via Samba or FTP or WebDAV?

You have selected mounting by UUID /dev/sda2. This won’t work as the UUID looks like XXXXXXXX-XXX… Normally the UUID gets autofilled when you create the mount point.

I reattached the SSD to the device and now I have these mounts:

Is this the way that that I want this? What if I want to mount it to /mnt/foobar? Do I need to delete the current mounts and create a new one (or two)? Why is one physical devices split into two (sda1 and sda2)?

Sorry for the n00b questions.

sda1 seems kinda useless with 200 MB but for sda2 i would create a new mount point. Please post the add mountpoint screen so i can check it.

OK, I added it manually using the UI (clicked the Add button):

The two /tmp/run/mountd/sda* entries were removed and replace with the /mnt/jetdrive entry.

Which appears to have worked as desired:

root@turris:/mnt/jetdrive# df -h
Filesystem                Size      Used Available Use% Mounted on
/dev/mmcblk0p1            7.3G    166.1M      7.1G   2% /
tmpfs                   503.7M      1.5M    502.2M   0% /tmp
tmpfs                   512.0K      4.0K    508.0K   1% /dev
/dev/sda2               458.2G    955.9M    434.0G   0% /mnt/jetdrive

Will these settings remain if I reboot?

reboot and check if df -h lists the same :smiley: But they should

It worked. Thanks a lot for the help.