01_Process_Management.md

01. Process Management

---

⚙️ Foreground vs. Background Processes

By default, a command you run in a script executes in the foreground — the shell waits for it to finish before continuing. Appending & sends it to the background, allowing the script to continue immediately.

# Foreground: script waits here until sleep finishes
sleep 30
echo "30 seconds later..."

# Background: both commands start immediately
sleep 30 &
echo "This prints right away while sleep runs in background"

---

🔢 Job Control

Command	Meaning
command &	Run command in the background
jobs	List current background jobs
fg %1	Bring job #1 to the foreground
bg %1	Resume suspended job #1 in the background
Ctrl+Z	Suspend foreground job
$!	PID of the last background command

#!/usr/bin/env bash
# Parallel downloads
wget https://example.com/file1.zip &
PID1=$!
wget https://example.com/file2.zip &
PID2=$!

echo "Downloading file1 (PID: $PID1) and file2 (PID: $PID2)..."

# Wait for both to complete
wait $PID1 && echo "file1 downloaded."
wait $PID2 && echo "file2 downloaded."
echo "All downloads complete."

wait — Synchronize with Background Jobs

for server in web1 web2 web3; do
    ssh "$server" "sudo systemctl restart nginx" &
done

wait    # Wait for ALL background jobs to complete
echo "All servers restarted."

---

🔍 Inspecting Processes — ps, pgrep, pidof

# Show all running processes
ps aux
ps -ef

# Find processes by name
pgrep nginx              # Returns PIDs of all 'nginx' processes
pgrep -l nginx           # PID + name
pgrep -a nginx           # PID + full command line
pidof nginx              # Similar, slightly different behavior

# Filter with grep
ps aux | grep "[n]ginx"  # The [n] trick avoids matching the grep itself

---

💀 Killing Processes — kill, pkill, killall

# Send a signal to a specific PID
kill 1234           # SIGTERM (15) — graceful shutdown (default)
kill -9 1234        # SIGKILL — force kill (cannot be caught)
kill -HUP 1234      # SIGHUP (1) — reload config (graceful restart)

# Kill by name
pkill nginx         # Send SIGTERM to all 'nginx' processes
pkill -9 -u john    # Kill all processes owned by user 'john'
killall -HUP nginx  # Send SIGHUP to all processes named 'nginx'

Common Signals

Signal	Number	Meaning
SIGHUP	1	Hangup / Reload configuration
SIGINT	2	Keyboard interrupt (Ctrl+C)
SIGTERM	15	Graceful termination (default kill)
SIGKILL	9	Force kill — cannot be caught or ignored
SIGUSR1/2	10/12	User-defined signals

---

🪤 trap — Catching Signals and Cleanup

trap lets your script intercept signals and run cleanup code before exiting. This is essential for production scripts.

#!/usr/bin/env bash
TMPFILE=$(mktemp)
LOCKFILE="/var/run/myscript.lock"

# Cleanup function
cleanup() {
    echo "Script interrupted. Cleaning up..."
    rm -f "$TMPFILE" "$LOCKFILE"
    exit 1
}

# Run cleanup on these signals
trap cleanup SIGINT SIGTERM

# Create a lock file to prevent duplicate runs
if [[ -f "$LOCKFILE" ]]; then
    echo "Error: Script is already running (PID: $(cat $LOCKFILE))"
    exit 1
fi
echo $$ > "$LOCKFILE"

# --- Main script logic here ---
echo "Working... (PID: $$)"
echo "Press Ctrl+C to test cleanup."
sleep 60

# Normal cleanup at the end
rm -f "$TMPFILE" "$LOCKFILE"
echo "Done."

trap on EXIT

The EXIT pseudo-signal is the most useful — it fires whenever the script exits, whether normally or due to an error:

#!/usr/bin/env bash
tmpdir=$(mktemp -d)

# This will ALWAYS run, no matter how the script exits
trap 'rm -rf "$tmpdir"' EXIT

# Now we can use $tmpdir freely — it's automatically cleaned up
cp important_files/* "$tmpdir/"
process_files "$tmpdir"

---

🏆 Summary

Tool	Purpose
& and wait	Run tasks in parallel; wait for completion
$!	Get the PID of the last background process
ps aux	View all running processes
pgrep name	Find PIDs by process name
kill -SIGNAL pid	Send a signal to a process
pkill name	Kill processes by name
trap cmd SIGNAL	Run code when a signal is received

👉 Next: Scheduling with Cron