🛑 Extended Traps and Signals

Signals are asynchronous notifications sent to processes. Traps let you intercept and handle them gracefully.

🧭 Common Signals

Signal	Number	Description	Default Action
SIGINT	2	Ctrl+C — Interrupt	Terminate
SIGTERM	15	Termination request	Terminate
SIGKILL	9	Force kill (cannot trap)	Terminate
SIGHUP	1	Hangup (terminal closed)	Terminate
SIGSTOP	19	Stop process (cannot trap)	Stop
SIGCONT	18	Continue stopped process	Continue
SIGUSR1	10	User-defined 1	Terminate
SIGUSR2	12	User-defined 2	Terminate

---

🧪 Basic Trap Usage

Catch Ctrl+C

trap 'echo "Interrupted!"; exit 1' INT

while true; do
    sleep 1
    echo "Running..."
done

Press Ctrl+C → prints "Interrupted!" and exits cleanly.

Multiple Signals

cleanup() {
    echo "Cleaning up..."
    rm -f /tmp/myapp.lock
    exit 0
}

trap cleanup INT TERM EXIT

This catches: - Ctrl+C (INT) - kill command (TERM) - Any exit (EXIT)

---

🧠 Ignoring Signals

Prevent accidental termination:

trap '' INT    # Ignore Ctrl+C
trap '' TERM   # Ignore termination requests

Useful for critical sections:

critical_section() {
    trap '' INT TERM   # Protect during critical work

    # Sensitive operations
    write_to_database

    trap - INT TERM    # Restore default handling
}

---

🧪 Resetting Traps

Restore default behavior:

trap - INT    # Reset INT to default
trap -        # Reset all traps

Check current trap:

trap -p INT   # Shows what INT trap is set to

---

🧠 Practical Use Cases

Graceful Shutdown

shutdown_requested=false

graceful_shutdown() {
    echo "Shutdown signal received..."
    shutdown_requested=true
}

trap graceful_shutdown TERM INT

while [ "$shutdown_requested" = false ]; do
    process_next_job
    sleep 1
done

echo "Performing final cleanup..."
cleanup_resources
exit 0

Temporary Signal Handlers

pause_processing() {
    trap 'echo "Paused. Resume with fg."' TSTP
    kill -STOP $$   # Pause this process
}

resume_processing() {
    trap - TSTP
    kill -CONT $$   # Resume this process
}

---

🧪 Sending Signals to Processes

# Find process by name
pid=$(pgrep nginx | head -1)

# Send signals
kill -TERM $pid   # Graceful termination
kill -KILL $pid   # Force kill (last resort)
kill -HUP $pid    # Reload config (common for daemons)
kill -USR1 $pid   # Custom action (e.g., rotate logs)

Check if process responded:

kill -0 $pid 2>/dev/null && echo "Process exists" || echo "Process gone"

---

🧠 Signal Handling in Functions

Traps are global — they affect the entire shell:

setup_trap() {
    trap 'cleanup' EXIT
}

cleanup() {
    echo "Cleaning up from $FUNCNAME"
}

setup_trap
# cleanup will run when script exits, even from nested functions

---

🧪 Debugging Signal Issues

Trace signal delivery:

trap 'echo "Received signal: $1"' DEBUG

Or use strace:

strace -e trace=signal ./myscript.sh

---

🧾 Summary

• Trap INT, TERM, EXIT for graceful shutdowns.
• Use trap '' to protect critical sections.
• SIGKILL and SIGSTOP cannot be trapped.
• Send custom signals with kill -USR1, etc.
• Always clean up resources in trap handlers.
• Test signal handling under load.

👉 Continue to: Subprocesses and Subshells