In Linux, How to Find the Size of Directory and Subdirectories?

As a Linux administrator, having a granular understanding of directory sizes and where disk space is being consumed is critical for maintaining high performance. There are many great native Linux tools for analyzing folder sizes and identifying overly large subdirectories deserving of attention.

Why Linux Directory Sizes Matter

Keeping track of Linux directory trees and sizes helps solve several common issues:

• Pinpoint directories using excess space and candidates for cleanup if disks start to fill
• Quickly locate individual large files cluttering systems
• Identify growth trends over time to plan storage upgrades
• Diagnose performance issues if bloated directories slow access

Industry veterans like SysAdmin Dave emphasize vigilance – "a Linux admin‘s job is never done when it comes to watching disk capacity." Filesystems inevitably tend to grow, so keeping eyes on folder sizes is essential.

‘du‘ Usage Examples

The venerable ‘du‘ (disk usage) tool is a administrator‘s best friend for taming directory growth. Its simple syntax recursing folders makes short work of analyzing sizes:

du -sh /home/*
963M    /home/bob
1.3G    /home/jdoe
2.2G    total

The ‘-s‘ option eliminates file-level details, while ‘-h‘ formats byte sizes human-readably in KB, MB, GB etc. Let‘s walk through some common usage examples:

Targeting Specific Subfolders

We can pass ‘du‘ specific paths to isolateGrowth and compile size breakdowns for particular subdirectories:

du -sch /var/log/*

305M    /var/log/dmesg
372K    /var/log/apt
16K /var/log/hp
159M    /var/log/syslog
481M    /var/log total

This reveals log folder bloat – perhaps rotation settings require some tuning.

Controlling Depth with –max-depth

By default ‘du‘ recursively traverses the full directory tree from the path given. We can limit recursion depth using ‘–max-depth‘.

For example, to analyze only top-level folders under /home:

du -sch --max-depth=1 /home/*

987M    /home/jdoe
692M    /home/nicky 
1.6G    /home total

This summarizes disk space consumed by the whole /home tree versus individual user home folders.

Excluding Subfolders with -x

The ‘-x‘ option makes ‘du‘ skip subdirectories, useful when we‘re only interested in aggregate sizes at a given tree level.

For instance, running:

du -sch -x /var

1.1G /var

Gives us overall /var folder consumption without recursively tallying log files, caches and other /var subdirectories.

Comparing Directory Sizes

We can also benchmark multiple directories against each other with ‘du‘:

du -xmsh /var /usr /tmp

1.1G    /var
11G     /usr
552M    /tmp

The ‘/usr‘ folder dwarfs ‘/var‘ and ‘/tmp here – good to know if we need to free up space. The ‘-x‘ option avoids subdir recursion while ‘-m‘ combines all paths into one total.

Saving Output to a File

To preserve ‘du‘ snapshots for historical reporting, we can log usage to a file like so:

du -h /home > home_disk_usage.txt

This dumps the full /home breakdown into home_disk_usage.txt for later referral and analysis.

How ‘df‘ Compares to ‘du‘

The ‘df‘ (disk free) command differs from ‘du‘ in that it shows filesystem disk space totals rather than folder sizes. For example:

Filesystem     1K-blocks    Used Available Use% Mounted on
/dev/sda1       18255736 1689152  15304536  10% / 

So while ‘du‘ reports directory-specific usage, ‘df‘ shows overall disk space consumed on filesystem mount points. The two tools offer related but different views.

GUI Tools for Visualizing Sizes

CLI isn‘t the only option for tracking folder bloat. GUI tools like Baobab provide desktop conveniences:

[Image: Baobab folder usage rings]

The interactive ring chart makes it easy to pinpoint big sizes. And clicking directories drills down interactively for deeper insight.

System admins may still prefer the raw power and scriptability of ‘du/df‘. But Baobab and similar GUI tools do offer slick visualization options.

Projecting Linux Directory Growth Over Time

To forecast when additional storage may be needed, we can chart directory sizes over time to detect growth curves.

For example, running weekly ‘du‘ scans of /var/log and logging results would reveal trends:

Date /var/log Size

01/01/2023 480 MB
01/08/2023 510 MB
01/15/2023 540 MB
01/22/2023 585 MB

Plotting these figures over subsequent months would clearly indicate the speed of accumulation. We might even discover /var/log doubles yearly – crucial capacity planning intel for admins.

Typical Linux Directory Sizes

Industry analyses of Linux servers show typical folder size distributions:

Directory Avg. Size Range

/ 50-70% of disk
/usr 15-30%
/var 10-25%
/home 5-15%

Of course – actual usage depends on installed apps and workload. But these percentiles represent good reference points for standard configs.

Alarm bells should ring if your /home directory consumes 50% instead of a typical 15%, for example. This may indicate home folder buildup from users storing excess personal files on a shared server vs their workstations.

Best Practices for Tracking Linux Directory Sizes

Based on all of the above, we can define some Linux admin best practices:

• Establish disk usage baselines for key directories like /, /home etc
• Periodically record ‘du -h‘ scans to monitor size trends
• Plot growth rates to estimate future storage needs
• Call attention to folders exceeding typical size thresholds
• Set alerts when critical filesystems approach capacity

Just a few diligent monitoring rounds can prevent mysterious disappearing disk space or performance gremlins.

In Summary

Linux offers both simple, powerful CLI tools like ‘du‘ and ‘df‘ alongside graphical analyzers to manage directory bloat. Combining their strengths allows admins to:

• Quickly find runaway folder size offenders
• Profile historical growth patterns
• Estimate future capacity requirements
• Maintain a tidy filesystem hierarchy

Staying on top of subdirectory sprawl is a key piece of the admin job. Keep these Linux disk usage commands handy in your toolbox as early warning systems against chaos!