Shell Scripting — Process Management

# ── Run in background ─────────────────────────────────────
sleep 10 &                        # run in background, get PID
BG_PID=$!                         # PID of last background job
echo "Started PID ${BG_PID}"

# ── wait — block until background job finishes ────────────
wait "${BG_PID}"                   # wait for specific PID
echo "Exit code: $?"

wait                               # wait for ALL background jobs

# ── Parallel execution — run multiple tasks at once ───────
pids=()

for server in web-01 web-02 web-03; do
  echo "  Deploying to ${server}..."
  ssh "${server}" "./deploy.sh" &
  pids+=($!)
done

# Wait for all and check exit codes
failed=0
for pid in "${pids[@]}"; do
  wait "${pid}" || (( failed++ ))
done
(( failed > 0 )) && echo "ERROR: ${failed} deployments failed" >&2

# ── jobs — list background jobs ───────────────────────────
jobs                               # show background jobs
jobs -l                            # with PIDs
fg %1                             # bring job 1 to foreground
bg %1                             # send job 1 to background

# ── Common signals ────────────────────────────────────────
# SIGINT  (2)  — Ctrl+C        — interrupt
# SIGTERM (15) — kill pid      — polite termination
# SIGKILL (9)  — kill -9 pid  — force kill (cannot be trapped)
# SIGHUP  (1)  — terminal close / reload config
# EXIT         — pseudo-signal, runs when script exits for any reason

# ── trap syntax ───────────────────────────────────────────
# trap 'command' SIGNAL [SIGNAL ...]

# ── Cleanup on any exit ───────────────────────────────────
TMPFILE=$(mktemp)
LOCKFILE=/tmp/myapp.lock

cleanup() {
  echo "  Cleaning up..."
  rm -f "${TMPFILE}" "${LOCKFILE}"
  kill 0 2>/dev/null   # kill all child processes in group
}
trap cleanup EXIT          # runs on ANY exit
trap cleanup INT TERM HUP  # also on signals

# ── Graceful shutdown with message ────────────────────────
handle_sigterm() {
  echo "  SIGTERM received — shutting down gracefully..."
  # finish current work, then exit cleanly
  SHUTDOWN=true
}
trap handle_sigterm TERM

# Main loop checks the flag
SHUTDOWN=false
while ! "${SHUTDOWN}"; do
  process_one_job
done
echo "  Shutdown complete"

# ── ps — process status ───────────────────────────────────
ps aux                            # all processes, detailed
ps aux | grep nginx               # filter by name
ps -p 1234                        # specific PID
ps -u vriddh                      # processes of user
ps --sort=-%cpu | head -10        # top CPU consumers
ps --sort=-%mem | head -10        # top memory consumers

# ── pgrep / pkill — find/kill by name ────────────────────
pgrep nginx                       # get PIDs of nginx processes
pgrep -l nginx                    # with process names
pgrep -u vriddh                   # processes of user
pgrep -x nginx                    # exact name match

pkill nginx                       # kill all nginx processes (SIGTERM)
pkill -9 myapp                    # force kill
pkill -HUP nginx                  # send SIGHUP (reload)
pkill -u vriddh sleep             # kill vriddh's sleep processes

# ── kill — by PID ─────────────────────────────────────────
kill 1234                         # SIGTERM (15)
kill -9 1234                      # SIGKILL (9) — last resort
kill -HUP 1234                    # SIGHUP — reload config
kill -0 1234                      # check if PID exists (no signal sent)

# ── Check if process is running ──────────────────────────
is_running() {
  local name="${1}"
  pgrep -x "${name}" >/dev/null 2>&1
}

if is_running "nginx"; then
  echo "  ✔ nginx is running"
else
  echo "  ✘ nginx is NOT running"
fi

# ── nohup — survive terminal close ───────────────────────
nohup ./long_process.sh &         # output goes to nohup.out
nohup ./long_process.sh > /var/log/process.log 2>&1 &
echo "PID: $!"

# ── disown — detach background job from shell ─────────────
./long_process.sh &
disown                            # detach last background job
disown %1                         # detach job 1
disown -h %1                      # keep in jobs list but ignore SIGHUP

# ── Lockfile — prevent duplicate script runs ──────────────
LOCKFILE="/tmp/$(basename "$0").lock"

acquire_lock() {
  if [ -f "${LOCKFILE}" ]; then
    local locked_pid
    locked_pid=$(cat "${LOCKFILE}" 2>/dev/null)
    if kill -0 "${locked_pid}" 2>/dev/null; then
      echo "ERROR: Already running (PID ${locked_pid})" >&2
      exit 1
    fi
    echo "  Stale lock found — cleaning up"
  fi
  echo $$ > "${LOCKFILE}"
}
trap 'rm -f "${LOCKFILE}"' EXIT

acquire_lock

# ── flock — kernel-level locking (more reliable) ──────────
# Exclusive lock — only one instance at a time
exec 9>/tmp/myapp.lock
flock -n 9 || { echo "Already running" >&2; exit 1; }

# Or one-liner with flock
flock -n /tmp/myapp.lock ./myapp.sh || echo "Already running"