Shell Scripting for SysAdmins: From Basics to Automation

I see shell scripting everywhere in my work, from managing cloud infrastructure to automating simple daily tasks.

For many, shell scripting is the bridge between just “using” a system and actually “managing” it. It is how you move from manual tasks to reliable, repeatable automation.

Here is my practical take on what every sysadmin should know.

Part 1: The Shebang and The Basics

A shell script is just a text file with commands. The key is the first line—the Shebang—which tells the system how to run it.

#!/bin/bash

Once written, don’t forget the execution bit: chmod +x script.sh

Part 2: Variables and Input

In Bash, everything is a string by default. Keep it simple.

Variables: name="Alice"
Access: echo "Hello, $name"
Arguments: $1, $2 refer to command-line inputs. $0 is your script name.

Avoid complexity. If you are doing complex types, you probably need a different language.

Part 3: Control structures

Real automation makes decisions. Use if and loop constructs to handle system state.

Directory check (Very common)

if [ -d "$dir" ]; then
    echo "Directory exists."
else
    echo "Directory does not exist."
fi

Simple Loops

for i in {1..5}; do
    echo "Iteration $i"
done

Part 4: Building Robust Scripts

As you scale from one-off tasks to managed automation, standardizing your approach is crucial.

1) Use functions

Group code to keep your scripts clean and maintainable.

2) Error handling is not optional

Check the exit status of every critical command. $? is your best friend.

mkdir /some/data
if [ $? -eq 0 ]; then
    echo "Success."
else
    echo "Failed. Check permissions."
fi

If it failed, fail fast. Don’t let the script continue in a broken state.

Part 5: Automation with Cron

The goal is to stop running scripts manually. cron is the standard tool for scheduled tasks.

crontab -e lets you set the timing:

0 3 * * * /path/to/script.sh

(Runs at 3 AM every day)

NetScaler and Infrastructure Note

When you are automating infrastructure—like I do with NetScaler or cloud APIs—treat your scripts like production code:

Fail closed: If a script can’t confirm state, stop.
Validate input: Never assume environment variables are set correctly.
Log everything: If it fails at 3 AM, you need to know why when you check the logs at 8 AM.

Final takeaway

Shell scripting isn’t about being a “developer.” It’s about being a “sysadmin” who values their time.

Master the basics: variables, conditionals, loops, and error checking. If you automate your manual work, you free yourself to focus on architecture and strategy instead of maintenance.