What are the essential Bash scripting concepts an SRE should know, and how do you write a reliable script for production use?

Question

Accepted Answer

Bash scripting is the most fundamental automation skill for an SRE because it is available on virtually every Linux system without installing anything extra. Whether you are automating log cleanup, health checks, or deployment tasks, Bash is often the fastest path from manual task to automated solution.

Every production Bash script should start with a proper shebang line (#!/bin/bash) and strict mode settings. The set -euo pipefail combination is essential: -e makes the script exit immediately if any command fails, -u treats unset variables as errors instead of silently expanding to empty strings, and -o pipefail makes pipeline failures propagate correctly. Without these settings, scripts silently continue after errors, which can cause cascading damage.

Variables and string handling are the foundation. Use meaningful variable names and always quote variable expansions ("$variable" not $variable) to handle filenames with spaces correctly. Use curly braces for clarity: "${variable}_suffix". Read-only variables should be declared with readonly to prevent accidental modification.

Conditionals use the if-then-else-fi structure or the more concise [[ ]] test syntax. Common tests include checking if a file exists (-f), if a directory exists (-d), if a string is empty (-z), or comparing numbers (-eq, -lt, -gt). For SRE work, you frequently check command exit codes: if ! command; then echo "command failed"; exit 1; fi.

Loops process multiple items. For loops iterate over lists: for server in web1 web2 web3; do ssh "$server" uptime; done. While loops read input line by line: while read -r line; do process "$line"; done < input.txt. Always use -r with read to prevent backslash interpretation.

Functions keep scripts organized and reusable. Define them with function_name() { ... } and call them like commands. Functions can return exit codes and produce output that can be captured with command substitution: result=$(my_function).

Error handling and logging make scripts production-ready. Create a log function that timestamps output: log() { echo "[$(date '+%Y-%m-%d %H:%M:%S')] $*"; }. Use trap to clean up temporary files on exit: trap 'rm -f "$tmpfile"' EXIT. For critical operations, add explicit error checks with informative messages.

Command-line arguments make scripts flexible. Use getopts for option parsing or simple positional parameters ($1, $2) with validation. Always validate inputs and provide a usage message when arguments are missing or invalid. This prevents scripts from doing unexpected things when called incorrectly.

For anything beyond 100-200 lines of Bash, consider switching to Python. Bash excels at orchestrating other commands and simple text processing, but complex logic, data structures, and error handling are significantly easier in Python.