-
Notifications
You must be signed in to change notification settings - Fork 1.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
map_server process will not exit when ctrl-c #1163
Comments
I tried replicating, and I see this issue too.
If I wait ~4 minutes, all nodes shutdown. |
Are you sure they're not actually shut down but some DDS "stuff" hasn't updated that its gone? If you call a service from one of these guys, does it respond? Are they publishing topics? |
No, it just shows up, but I believe it is not really running. I cannot echo topics. |
If |
I think this may be DDS related. Since the node is killed, it can't tell its subscribers that it is dead. The other nodes need to discover that this node is actually dead (not responding). The active data readers and writers send heartbeats and acknowledgments to make sure that they are still alive. It may take some time to drop a dead publisher or a subscriber from the list. Same for ROS daemon. Since it listens to all the discovery protocol messages between data writers and readers, it needs some time to update its node list cache. You can run PS: I recently started learning about DDS and how it works, so I might be wrong. |
Well something is still running @mlherd if |
@SteveMacenski I should have used the quote reply. My intent was replying to @yathartha3's comment:
I will debug this issue. |
Hey! I am also seeing this with opensplice on 0.2.4 It doesn't reproduce consistently, but when it does |
mhm, next time try |
I have also tried |
@Jconn I can confirm I've seen map_server persist in the past once its launch file has been killed. Never figured out why. |
It doesn't reproduce consistently. This problem has not occurred in my recent use. When I ctrl+c the launch file, the map_server will shutdown together.
|
I couldn't reproduce this issue. I tested it with the most recent Nav2 Dashing release. map_server process always dies after I kill the launch. However, I have seen this issue happening before. After I kill the launch file, sometimes map_server node will still be listed when I run ros2 topic list, but /map topic is not being published. If I run ros2 daemon stop, map_server node disappears. |
I have tried it many times, but I can't reproduce this issue. 1- launch Navigation2
5-
6-
7- 8-
I will close this issue for now, but please feel free to reopen it, if you still face this issue. |
The problem still persists, anyone have solution? I use dashing |
Bug report
Required Info:
Steps to reproduce issue
Expected behavior
All nodes will shutdown.
Actual behavior
The /map topic is still published
ros@neuron:~$ ps aux | grep ros2
ros 9379 1.4 0.6 1938704 99576 ? Sl 15:28 0:34 /home/ros/yunji_ros2/install/nav2_map_server/lib/nav2_map_server/map_server __node:=map_server __params:=/tmp/tmp2p9l3cr5
ros 19342 0.0 0.0 21908 1112 pts/3 S+ 16:10 0:00 grep --color=auto ros2
ros 19792 0.4 0.6 738872 101764 ? Sl 14:17 0:31 gedit /home/ros/yunji_ros2/src/robot_bringup/launch/robot_bringup.launch.py
ros 22430 0.7 0.4 1933116 80928 ? Sl 15:38 0:13 /home/ros/yunji_ros2/install/nav2_map_server/lib/nav2_map_server/map_server __node:=map_server __params:=/tmp/tmpo7_bh8g4
Additional information
Feature request
Feature description
Implementation considerations
The text was updated successfully, but these errors were encountered: