Roughly a week after installing my 2nd ET9000 I powered on the box to be presented with a blank green screen with a spinning Vix logo in the top left hand corner (the VFD was stuck at “booting…”).

Fortunately I was able to SSH into the box and access the crash log:

IOError: [Errno 28] No space left on device: '/usr/share/enigma2/ViX_HD/skin.xml.tmp'
---- saving lame channel db
FATAL: couldn't save lame channel db!

There were hundreds of crash logs with date/time stamps just minutes apart (suggesting the box was stuck looping through the boot routine). Each of the logs had a number of entries suggesting the root cause was a full disk.

df -h confirmed this to be the case:

Filesystem Size Used Available Use% Mounted on
/dev/root 122.0M 121.5M 552.0K 100% /
tmpfs 64.0K 8.0K 56.0K 13% /dev
/dev/sda2 15.4G 67.4M 15.3G 0% /media/usb
tmpfs 155.4M 804.0K 154.6M 1% /var/volatile
tmpfs 155.4M 0 155.4M 0% /dev/shm

But why was /dev/root full when my recordings/crash logs are all stored in /media/hdd which points to /media/net/tv?

I ran a ton of linux commands to try and identify large files/directories but nothing stood out- I was almost ready to re-flash the Vix firmware, then a had a brainwave…

If at some point the box had been unable to mount /media/net/tv (I had suffered a power cut recently) then where would recordings & crash logs be kept? Again- I couldn’t find the location anywhere, so I thought I’d force it…

I took the samba share offline and restarted the box- voila! /media/hdd was mounted but this time I couldn’t see the normal recordings and crash logs- just a single recording…

I ran rm * then df –h

/dev/root 122.0M 60.2M 61.8M 49% /
tmpfs 64.0K 8.0K 56.0K 13% /dev
tmpfs 155.4M 688.0K 154.7M 0% /var/volatile
tmpfs 155.4M 0 155.4M 0% /dev/shm

The ET9000 flash had 51% free space again! I placed the samba share back online, rebooted the box and everything is back to normal.

Thanks to the community @ for helping me troubleshoot/solve the problem!