Closed VRichardJP closed 2 years ago
This is just a quick thought, but maybe the zombie process is lingering (I encounter this issue from time to time). If the zombie lanelet2_loader that loaded yyy map lingered, the above issue may happen. You can just check it by ros2 node list
without launching anything.
If there exists some zombie processes, killing all the ROS processes may solve the issue. e.g.:
pkill ros2
pkill rviz2
pkill aggregator_node
ps aux | grep python3 | grep ros2 | grep -v grep | awk '{ print "kill ", $2 }' | sh
ps aux | grep python3 | grep rqt_reconfigure | grep -v grep | awk '{ print "kill ", $2 }' | sh
ps aux | grep component_container | grep -v grep | awk '{ print "kill ", $2 }' | sh
ps aux | grep robot_state_publisher | grep -v grep | awk '{ print "kill ", $2 }' | sh
ps aux | grep topic_tools/relay | grep -v grep | awk '{ print "kill ", $2 }' | sh
ps aux | grep "ros-args" | grep -v grep | awk '{ print "kill ", $2 }' | sh
You are right, I can see many nodes in ros2 node list
. There are not because of some zombie process, but because there is another instance of autoware running on the same network. Next time I will be careful and start ROS2 with localhost only ;)
Checklist
Description
I am sorry about the confusing title, as I am myself confused over the following case:
I start autoware planning simulator with:
In the first few lines of logs, I can see the lanelet2_map_loader has crashed with the message:
Suprisingly the file exist:
More surprisingly, I see in Rviz that another lanelet and pcd have been loaded! (located in /data/autoware_data_universe/yyy)
Just to double check, this is what my xxx map looks like:
Expected behavior
The correct map is loaded
Actual behavior
Magic
Steps to reproduce
?
Versions
No response
Possible causes
No response
Additional context
No response