How to Check Disk Space in Linux

The df command reports the amount of disk space used and available on mounted filesystems.

df

For most troubleshooting tasks, use the human-readable option:

df -h

The -h option displays sizes in a more readable format such as GB and MB.

df -h

A typical result contains columns such as:

• Filesystem — the filesystem or device.
• Size — total filesystem size.
• Used — space currently used.
• Avail — available space.
• Use% — percentage of space used.
• Mounted on — filesystem mount point.

The most important column during a disk-space incident is usually Use%.

You can check the filesystem containing a particular path.

df -h /var

This is useful when an application reports that a directory such as /var is running out of space.

While df tells you about filesystem capacity, du helps identify how much space files and directories are using.

Check the size of a directory:

du -sh /var/log

The options mean:

• -s — show a summary.
• -h — use human-readable sizes.

A useful first step is to inspect the major directories under a filesystem.

sudo du -xh --max-depth=1 /

For a particular directory:

sudo du -xh --max-depth=1 /var

Sort the results to make large directories easier to identify:

sudo du -xh --max-depth=1 /var | sort -h

Once you identify a directory consuming significant space, use find to locate large files.

For example, find files larger than 100 MB:

sudo find /var -type f -size +100M -print

To find files larger than 1 GB:

sudo find /var -type f -size +1G -print

You can combine find with other Linux commands to identify large files.

sudo find /var -type f -printf '%s %p\n' 2>/dev/null | sort -nr | head -20

This approach lists the largest files first.

Command behavior can vary slightly between Linux implementations, so verify the supported find options on your distribution with:

man find

The lsblk command displays block devices and their relationships.

lsblk

For filesystem information:

lsblk -f

This is useful when you need to understand which disks, partitions and filesystems are present on a Linux system.

A filesystem can have available storage but still fail to create new files if it has exhausted its available inodes.

Check inode usage with:

df -i

If inode usage is at or near 100%, look for directories containing very large numbers of small files.

It is possible for df and du to report different amounts of used space.

One common reason is that a process can keep an already deleted file open. The directory entry is removed, but the filesystem space can remain allocated until the process releases the file.

When investigating this situation, check for open deleted files using tools available on your distribution, such as:

sudo lsof +L1

If lsof is installed, this can help identify processes holding deleted files open.

Logs are a common source of unexpected disk consumption on servers.

Start by checking the size of common log locations:

sudo du -sh /var/log
sudo du -xh --max-depth=1 /var/log | sort -h

Then inspect particularly large files:

sudo find /var/log -type f -size +100M -ls

Before deleting logs, determine whether log rotation or the application generating the logs needs attention.

Temporary files, application caches, package data and container storage can also consume significant disk space.

Inspect directories rather than deleting their contents blindly:

sudo du -xh --max-depth=1 /tmp | sort -h
sudo du -xh --max-depth=1 /var | sort -h

The exact directories worth investigating depend on the software and Linux distribution installed on the system.

On systems running Docker, images, containers, volumes and build cache can consume substantial disk space.

If Docker is installed, inspect its disk usage:

docker system df

This provides an overview of Docker's storage consumption.

When a Linux server reports low disk space, use a structured workflow instead of immediately deleting files.

1. Run df -h to identify the full filesystem.
2. Run df -i to check inode usage.
3. Use du to identify large directories.
4. Use find to locate unusually large files.
5. Inspect logs, application data, caches and backups.
6. Check for deleted files still held open by processes if df and du disagree.
7. Make the smallest safe cleanup or configuration change.
8. Run df -h again and validate the application.

Deleting files without identifying the cause

Removing a large file may temporarily create free space without fixing the process that caused the growth.

Using rm on system directories

Never delete system files simply because they appear large. Determine what they are and whether the associated service needs them.

Ignoring inode usage

A filesystem can have free capacity while running out of inodes.

Ignoring open deleted files

Deleting a log file doesn't necessarily immediately return all its storage if a running process still has the file open.

Cleaning production systems without a plan

Disk cleanup should be controlled, documented and validated, especially on production servers.

# Show filesystem usage
df -h

# Show inode usage
df -i

# Show filesystem containing a path
df -h /var

# Show directory size
du -sh /var/log

# Show immediate subdirectory sizes
du -xh --max-depth=1 /var | sort -h

# Find files larger than 100 MB
find /var -type f -size +100M -print

# Show disks and filesystems
lsblk -f

# Find deleted files still held open
sudo lsof +L1