Build a Simple Bash Script to Monitor CPU, Memory, and Disk Usage on Linux

Monitoring system resources is essential for server operations, especially when deploying new applications. This guide shows how to write a compact Bash script that reports memory, disk, and CPU usage as three columns.

1. Monitoring Memory

free -m | awk 'NR==2{printf "%.2f%%\t\t", $3*100/$2 }'

The free -m command displays total and used memory. Using awk we extract the second line, where $3 is used memory and $2 is total memory, then calculate the usage percentage.

2. Monitoring Disk

df -h | awk '$NF=="/"{printf "%s\t\t", $5}'

df -h

reports filesystem usage. The awk filter selects the line whose mount point is / and prints the fifth field, which contains the used‑percentage (e.g., 7%).

3. Monitoring CPU

top -bn1 | grep load | awk '{printf "%.2f%%\t\t
", $(NF-2)}'

top -bn1

runs top in batch mode for a single iteration. grep load extracts the line with the load averages, and awk prints the field two positions before the end, which corresponds to the current CPU load percentage.

4. Assembling the Script

#! /bin/bash
printf "Memory\t\tDisk\t\tCPU
"
end=$((SECONDS+3600))
while [ $SECONDS -lt $end ]; do
  MEMORY=$(free -m | awk 'NR==2{printf "%.2f%%\t\t", $3*100/$2 }')
  DISK=$(df -h | awk '$NF=="/"{printf "%s\t\t", $5}')
  CPU=$(top -bn1 | grep load | awk '{printf "%.2f%%\t\t
", $(NF-2)}')
  echo "$MEMORY$DISK$CPU"
  sleep 5
done

The script prints a header, then enters a while loop that runs for one hour ( 3600 seconds). Each iteration updates the three variables, echoes them as a tab‑separated line, and pauses five seconds.

5. Optional Logging

You can redirect the script output to a log file:

./stats.sh >> log.txt

6. Stress Testing (Optional)

To generate load for testing, install stress (e.g., on CentOS) and run a one‑hour stress test:

# yum install stress
# stress -c 2 -i 1 -m 1 --vm-bytes 128M -t 3600s

While the stress test runs, the monitoring script will show increasing CPU and memory percentages.

Running the script alongside the stress test produces output such as:

Memory  Disk  CPU
20.48%  7%    1.21%
20.48%  7%    1.02%
... (continues every 5 seconds)

This completes a practical Bash tutorial for system administrators to monitor key resource metrics continuously.