Monitoring Disk Usage

Keeping an eye on your server's disk usage is crucial. Imagine a website suddenly going down because its logs filled up the entire disk! Or a database slowing to a crawl due to heavy disk activity.

Monitoring helps you:

• Prevent outages: Catch full disks before they cause problems.
• Optimize performance: Identify disks with high I/O that might be bottlenecks.
• Plan capacity: Understand growth trends to add storage proactively.

The df command stands for 'disk free'. It's your go-to tool for checking the overall disk space usage of mounted file systems on your server.

It provides a summary of total space, used space, available space, and percentage used for each mounted partition.

Let's see df in action. The -h flag makes the output human-readable, showing sizes in KB, MB, or GB.

Try running this command to see your server's disk space at a glance:

The output of df -h typically has these columns:

• Filesystem: The name of the disk partition or device.
• Size: Total size of the filesystem.
• Used: Amount of space currently in use.
• Avail: Amount of space still available.
• Use%: Percentage of disk space used.
• Mounted on: The directory where the filesystem is mounted.

Pay close attention to the Use% column to spot nearly full disks!

While df shows overall filesystem usage, du (disk usage) calculates the space used by specific files or directories.

This is extremely helpful for finding out which directories are consuming the most space, allowing you to pinpoint the source of high disk usage.

To find the size of a directory, you can use du with the -s (summarize) and -h (human-readable) flags, followed by the path to the directory.

Let's check the size of the /var/log directory, which often contains large log files:

It's important to know when to use which:

• df: Reports space usage for entire filesystems. It reads data from the filesystem's metadata.
• du: Reports space usage for files and directories. It scans through directory contents to calculate size.

Sometimes, df and du might show different numbers due to deleted files still held open by processes. A reboot or restarting the process usually syncs them up.

iostat is part of the sysstat package and provides detailed input/output (I/O) statistics for your storage devices.

This tool helps you monitor disk activity, which is crucial for identifying performance bottlenecks. High I/O wait times can significantly slow down your server.

To get a snapshot of your disk's I/O performance, you can use iostat. The -x flag provides extended statistics, and adding a number (e.g., 2) will make it update every 2 seconds.

This command might require sysstat to be installed (e.g., sudo apt install sysstat on Debian/Ubuntu).

When using iostat -x, look for these key metrics:

• %util: Percentage of CPU time during which I/O requests were issued to the device. High values (near 100%) indicate a busy disk.
• r/s, w/s: Number of read/write requests per second.
• rkB/s, wkB/s: Number of kilobytes read/written per second.
• await: The average time (in milliseconds) for I/O requests issued to the device to be served. High values suggest a slow disk.

These metrics help you gauge disk health and performance.

You've learned about three essential tools for monitoring disk usage and performance. Now, let's test your understanding!

Great job! You've mastered the basics of monitoring your server's disk usage and I/O performance.

You learned about:

• df: For checking overall free disk space on filesystems.
• du: For identifying disk usage by specific files and directories.
• iostat: For monitoring disk I/O performance and identifying bottlenecks.

Regularly using these tools will help you keep your server running smoothly and prevent unexpected issues.