def shutdown_plugin(self): self._timer.stop() rospy.signal_shutdown("odin rqtclose: Plugin closing") rospy.sleep(0.5) # Give ROS time to clean up Launch rqt with a timeout:
import signal signal.signal(signal.SIGTERM, my_shutdown_handler) If that handler calls sys.exit() without cleaning up rqt , you’ll see rqtclose errors. Modify your rqt plugin’s shutdown_plugin() method: odin rqtclose
#!/bin/bash # odin – correct wrapper for rqt rqt_pid="" function cleanup if [[ -n "$rqt_pid" ]]; then kill -TERM "$rqt_pid" wait "$rqt_pid" echo "odin rqtclose: clean exit" fi def shutdown_plugin(self): self
trap cleanup EXIT rosrun rqt_gui rqt_gui & rqt_pid=$! wait $rqt_pid Example of a safe shutdown in a Python rqt plugin: Fix 1 – Correct the Odin Wrapper Replace
def _spin_ros(self): rospy.rostime.wallsleep(0) # Allow ROS callbacks
rosnode kill /rqt_gui_py_node_xxxx Then improve your shutdown logic to call rosnode kill on itself (not recommended) or fix the plugin. Fix 1 – Correct the Odin Wrapper Replace any kill -9 with kill -TERM and add a wait loop: