@jacobperron friendly ping

The error message makes it sound like it's a bug in rospy.

@MatthijsBurgh Can you provide simple steps to reproduce the error?

@jacobperron I will get back to you when I have been able to isolate the issue into a reproducible error.

The situation in which I experienced this error the first time, was in a node launching other nodes.

Now I also experience this in nodes with an action_server simple_client. In both situations it happens during stopping of a node.

Exception in thread /hero/local_planner/action_server/result:
Traceback (most recent call last):
  File "/usr/lib/python3.8/threading.py", line 932, in _bootstrap_inner
    self.run()
  File "/usr/lib/python3.8/threading.py", line 870, in run
    self._target(*self._args, **self._kwargs)
  File "/opt/ros/noetic/lib/python3/dist-packages/rospy/impl/tcpros_pubsub.py", line 185, in robust_connect_subscriber
    conn.receive_loop(receive_cb)	    
  File "/opt/ros/noetic/lib/python3/dist-packages/rospy/impl/tcpros_base.py", line 846, in receive_loop
    self.close()
  File "/opt/ros/noetic/lib/python3/dist-packages/rospy/impl/tcpros_base.py", line 858, in close
    self.socket.close()
AttributeError: 'NoneType' object has no attribute 'close'

This is most likely a race condition due to a previous self.close() where the socket is set to None here. Given the various places where self.close() is called in TCPROSTransport because of errors (like maybe (improper) shutdown in your case) without proper thread sync, I wouldn't be surprised if it crashes sometimes.

We also experienced different cases where either the socket or the buffers were called as None, and after going through the code I think this can only be stabilized by refactoring the code to be more thread safe.

The complete solution of this issue would require to rewrite the classes to be thread safe. Anyway, a opened a PR with a workaround here: #2233

@jacobperron What do you think about #2233? Or would like to go for some locks?

I've merged the workaround. Thanks @madmage!

Hi there,
The same error raised when I use a simple function rospy.wait_for_message. I called this function in a sub-thread when I met this error first time, I thought it's the multi-thread caused. Then I made a simple test (single nodes without multi-thread), but same error raised. Here is the test code:

import rospy
from sensor_msgs.msg import JointState

import rospy
from sensor_msgs.msg import JointState

def main():
    rospy.init_node('test', anonymous=True)
    while not rospy.is_shutdown():
        js = rospy.wait_for_message('/iiwa/joint_states', JointState, timeout=5)
        print(list(js.position))

if __name__ == '__main__':
    main()

As the workaround in #2233 , a multi-thread set needed, but how can do it in single node and single thread case?

Thanks in advance!