Disk Management in Ubuntu: How to Check and Free Up Space

Introduction / Why This Is Needed

Occupied disk space is one of the most common reasons for Ubuntu slowdowns and errors like "No space left on device". This guide will show you how to not just delete files randomly, but to precisely locate the "space hogs" and clean them up safely. You'll gain full control over your disk using the terminal.

Requirements / Preparation

• Ubuntu 20.04 or newer (the instructions also apply to other Debian-based distributions).
• Access to the terminal (Ctrl+Alt+T).
• sudo privileges to clean system directories and install utilities.
• Basic familiarity with the command line.

Step 1: Install Necessary Utilities

Modern Ubuntu versions already include everything needed for basic analysis (df, du), but for convenience, let's install two powerful tools:

• ncdu — an interactive text-based analyzer (similar to du, but with navigation and sorting).
• duf — a beautiful and informative alternative to df (shows usage by partition, Inodes, filesystem types).

# Update the package list
sudo apt update

# Install ncdu and duf
sudo apt install -y ncdu duf

What the command does: apt install downloads and installs programs from official repositories. The -y flag automatically confirms the installation.

Step 2: Check Overall Disk Usage

First, let's see which partition is full.

# Run duf — this is the main command for a quick overview
duf

You will see a table where:

• Size — the total partition size.
• Used — used space.
• Avail — available space.
• Use% — the fill percentage (this is what we're looking at!).

Alternative: the old reliable df -h (human-readable).

df -h

Important: Note the mount point (Mounted on) of the full partition. Usually it's / (root partition) or /home.

Step 3: Find the Largest Directories

Now we search for what specifically is taking up space on the problematic partition. For example, if / is full:

# Analyze the root partition. This might take a while!
sudo ncdu /

How to use ncdu:

1. After launching, you'll see a list of folders sorted by descending size.
2. Up/Down arrows — navigation.
3. Enter — enter a folder (analyze its contents).
4. d — delete the selected file/folder (be careful!).
5. q — quit.

Tip: Start by analyzing /home (your personal files) and /var (logs, cache). Most often, the "culprits" are:

• Browser caches (~/.cache/*).
• Junk in ~/Downloads.
• Old package versions or kernels (/var/cache/apt/, /boot).
• Shared BitTorrent clients (/home/user/Downloads).

💡 Tip: If ncdu doesn't find anomalies, check hidden folders (starting with .) in your home directory. They are visible in ncdu.

Step 4: Clean the APT Package Manager Cache

APT stores downloaded .deb packages in /var/cache/apt/archives. They can be safely removed after installing programs.

# Clean the APT cache (frees up to hundreds of MB)
sudo apt-get clean

# Or a gentler option — remove only obsolete packages
sudo apt-get autoclean

Difference: clean removes all cached packages. autoclean removes only those that can no longer be downloaded (obsolete).

Step 5: Remove Old Logs and Temporary Files

System logs can grow large. journalctl manages systemd logs.

# Clear logs older than 7 days (safe)
sudo journalctl --vacuum-time=7d

# Or clear if logs take up more than 500 MB
sudo journalctl --vacuum-size=500M

Temporary files: Cleaning /tmp is usually safe, but some programs may use it for current data.

# Delete contents of /tmp (not the directory itself!)
sudo rm -rf /tmp/*

⚠️ Important: Do not manually delete files in /var/log directly (rm /var/log/*.log). Use journalctl or logrotate (configured by default). Deleting "by hand" can break running processes that write to these files.

Step 6: Check the Result

After each cleanup, return to Step 2 and see if the picture has changed.

duf

If space hasn't been freed, go back to ncdu and look for other large directories. The issue might be in user data (videos, iso images, virtual machines).

Potential Issues

Additional Methods (If the Main Ones Didn't Help)

1. Find Large Files Older Than N Days

Locate and remove giant artifacts (e.g., old backups):

# Find files larger than 100MB older than 30 days in /home
find /home -type f -size +100M -mtime +30 -exec ls -lh {} \;

Deletion (only after checking!):

find /home -type f -size +100M -mtime +30 -delete

2. Analyze Inodes (if df shows free space but you can't write)

df -i

If IUse% is close to 100%, the problem is the number of files, not their size. Often caused by millions of small files in a cache (e.g., npm, docker). Clean the respective caches.

3. Remove Old Kernels (caution!)

# List installed kernels
dpkg --list | grep linux-image

# Remove old ones (except the current one! Identify current via `uname -r`)
sudo apt-get remove linux-image-5.4.0-XX-generic

Never remove the last/current kernel!

Verification

1. Run duf — the Use% percentage should decrease.
2. Try creating a large test file (fallocate -l 1G testfile) to confirm space is actually free.
3. Restart services that failed due to disk shortage (e.g., sudo systemctl restart postgresql) and check their logs.

Final Recommendations

• Regularity: Once a month, run ncdu /home and sudo ncdu /var.
• Monitoring: Install duf and run it when you suspect problems.
• Automation: Add a cron job (crontab -e) to clean the APT cache and journals, but only if you are sure it's safe.
• Backups: Before mass deletion (especially via find ... -delete), ensure important data won't be affected.