Closed ValerioMagnago closed 1 year ago
Yes this is a known issue, but I have ignored it so far since it was not a blocking factor for the current users. Have you been able to debug the problem?
No, I didn't analyze the problem, but looking at the code this line here seems suspicious.
net_cv_->wait(net_lock, [&net_fail] { return net_fail; });
If my understanding is correct, here we just wait for a network failure but we do not react to a node termination.
Side note. I see you are using a ros::AsyncSpinner
and you set the number of threads as the number of processor core (https://docs.ros.org/en/noetic/api/roscpp/html/classros_1_1AsyncSpinner.html#aa5b9b685b7eebebf18ed42697e209897) why?
It is not possible to kill the node cleanly. Calling
rosnode kill
brings the driver into a deadlock situation and it is not possible to kill cleanly the node.