Fix No Space Left on Device | Generated by AI

Home 2026.02

Question: The system complains “no space left on device”, which is the root cause of the ongoing issues (including GDM failing to start). After reaching a text console with Ctrl+Alt+F3 (or similar), logging in, and running commands like sudo systemctl restart gdm or sudo systemctl status gdm -l, the problem persists. How to fix the “no space left on device” error and recover the system?

Answer:

The “No space left on device” error is now the primary blocker — even if ownership/permissions are partially fixed, GDM (and many other services) cannot start because the disk (usually the root / partition) is completely full. This often happens after reckless chown operations combined with package reinstall attempts, log spam, or temporary files.

Step-by-step fix from text console (TTY / Ctrl+Alt+F3)

Login at the text console with your username (lzwjava) and password.
Check which partition is full and why ```bash df -h

Look at / (root) — if it shows 100% or very close, that’s the issue

See the biggest space users (run these one by one)

du -sh /* | sort -hr | head -15

Most common culprits after broken upgrades/chown:

du -sh /var /var/log /var/cache /tmp /home /usr

Common offenders on broken Ubuntu systems:
- `/var/log/journal/` — huge logs from repeated failed starts
- `/var/cache/apt/archives/` — old .deb packages
- `/tmp/` — stuck temporary files
- `/var/lib/systemd/coredump/` — crash dumps

3. **Free up space — aggressive but safe cleanup**
Run these commands carefully (they are non-destructive):

```bash
# Clear old package cache (usually frees 1–5 GB)
sudo apt clean
sudo apt autoclean
sudo rm -rf /var/cache/apt/archives/*.deb 2>/dev/null

# Truncate huge logs (safe)
sudo journalctl --vacuum-time=2weeks
sudo journalctl --vacuum-size=100M
sudo truncate -s 0 /var/log/*.log /var/log/*/*.log 2>/dev/null

# Remove old kernels if many are installed (frees ~500 MB each)
sudo apt autoremove --purge

# Clear /tmp completely (only safe if you're not running important temp jobs)
sudo rm -rf /tmp/*
sudo mkdir /tmp
sudo chmod 1777 /tmp

# If /var/log/journal is huge:
sudo rm -rf /var/log/journal/*
sudo systemctl restart systemd-journald

Check space again

df -h
free -h   # also check RAM/swap — low RAM can cause similar symptoms

You need at least 500 MB–1 GB free on / for GDM and basic services to start reliably.

Try restarting GDM again ```bash sudo systemctl restart gdm

If it still fails, look at the real error now

sudo systemctl status gdm -l –no-pager | tail -30

Or full journal since last boot

journalctl -b -u gdm –no-pager | tail -40

Copy any new specific errors (permission denied, cannot write file, etc.) if it still doesn't work.

6. **If space is freed but GDM still won't start**
Re-run the earlier reinstall commands (now that there's space):
```bash
sudo apt update
sudo apt install --reinstall gdm3 ubuntu-gnome-desktop gnome-session ubuntu-session
sudo dpkg-reconfigure gdm3
sudo systemctl restart gdm

Switch to lightdm as fallback (much lighter, needs less space)

sudo apt install lightdm
sudo dpkg-reconfigure lightdm   # choose lightdm
sudo reboot

If lightdm starts successfully, you get a graphical login again — then you can fix GDM later or keep lightdm.

If you cannot free enough space from TTY

Boot back into the Ubuntu live USB (as you did before):

Mount your root partition to /mnt

Then run similar cleanup commands:

sudo mount /dev/nvme0n1pX /mnt   # replace with your root partition
sudo rm -rf /mnt/var/log/journal/*
sudo rm -rf /mnt/var/cache/apt/archives/*
sudo journalctl --rotate   # if journald is accessible

Or just delete largest files you identify with du.

This is almost always recoverable once space is freed.

Prevention note

Avoid running broad recursive commands on /usr, /var, /etc. Use targeted fixes only (e.g. chown $USER:$USER ~/somefolder).

If you share the output of df -h and du -sh /* | sort -hr | head (or the biggest folders), I can tell you exactly which directories to clean next.

References:

Back Donate