Refactor WireGuard keepalive script: reduce initial sleep time, increase failure threshold, and improve server IP resolution logic

persistentkeepalive.sh

@@ -5,17 +5,18 @@ IFACE="${INTERFACE:-wg0}"
 
 # Wait for the WireGuard interface to be fully up
 echo "WireGuard keepalive monitor: waiting 15s for interface $IFACE to come up..."
-sleep 15
+sleep 5
 
 PERSISTENT_KEEP_ALIVE=$(grep PersistentKeepalive "$CONF" | awk '{print $3}')
 # Prefer the Endpoint IP; fall back to the Address-derived gateway
 WG_SERVER_IP=$(grep Endpoint "$CONF" | awk '{print $3}' | cut -d: -f1)
-if [ -z "$WG_SERVER_IP" ]; then
+PING=$(ping -c 1 -W 5 "$WG_SERVER_IP" | grep -E '1 packets transmitted, 1 packets received')
+if [ ! -z "$PING" ]; then
   WG_SERVER_IP="$(grep Address "$CONF" | awk '{print $3}' | cut -d. -f1-3).1"
 fi
 
 PING_INTERVAL=${PERSISTENT_KEEP_ALIVE:-25}
-MAX_FAILURES=3
+MAX_FAILURES=10
 fail_count=0
 
 echo "WireGuard keepalive monitor started (target: $WG_SERVER_IP, interval: ${PING_INTERVAL}s, threshold: $MAX_FAILURES)"
@@ -27,10 +28,8 @@ while true; do
     fail_count=$((fail_count + 1))
     echo "WireGuard keepalive ping failed ($fail_count/$MAX_FAILURES) to $WG_SERVER_IP via $IFACE"
     if [ "$fail_count" -ge "$MAX_FAILURES" ]; then
-      echo "WireGuard connection is stuck after $MAX_FAILURES consecutive failures — forcing container restart..."
-      # SIGKILL tini (PID 1) so the container exits with code 137 and Docker restarts it
-      kill -9 1
-      exit 1
+      echo "WireGuard connection is stuck after $MAX_FAILURES consecutive failures ... forcing container restart..."
+      exec kill -9 1
     fi
   fi
   sleep "$PING_INTERVAL"