Exit codes

Every command in Linux returns an exit code when it finishes. 0 means success. Any nonzero value means "something went wrong".

Checking the last exit code

some-command
echo $?           # 0 if success, nonzero if failure

$? is a special variable that holds the exit code of the most recent command. It changes with every command, so capture it into a variable if you need it later:

some-command
code=$?
# ... other things ...
if (( code != 0 )); then
  echo "earlier command failed with $code"
fi

exit N: end the script with a code

exit 0        # succeed
exit 1        # generic failure
exit 2        # convention: misuse (invalid arguments)

exit N ends the script and returns exit code N to whoever called it. Without an argument, exit returns the last command's exit code.

Why exit codes matter

Every automation tool you'll ever use (cron, systemd, CI pipelines, monitoring probes) decides "did this succeed?" by looking at the exit code. Getting them right is the difference between "backup succeeded" and a silently broken pipeline.

Common conventions

Standard exit codes worth recognizing:

• 0 success
• 1 general failure
• 2 misuse (bad arguments, missing config)
• 126 command found but not executable
• 127 command not found
• 130 killed by Ctrl+C (SIGINT)
• 137 killed by SIGKILL (often OOM)
• 143 killed by SIGTERM

You don't have to memorize them, but recognizing 127 as "not found" saves time when debugging.