ros2 dashing FastRTPS comms failing between two computers
I have three computers on my network. One computer is publishing, and the other two are subscribing.
One of my two subscribers is not seeing the messages being published.
I am using Fast RTPS for communication.
I think I found a clue as to what's wrong while using wireshark. I took a look at what was happening on my network, and the faulty computer is giving me a ton of messages like this one:
Source Destination Protocol Length Info
192.168.50.116 192.168.50.109 RTPS 110 INFO_DST, ACKNACK
192.168.50.116 is my subscriber, 192.168.50.109 is my publisher
I am ping-testing from my publisher to my faulty subscriber, and I'm getting ~700kB/s, so I'm skeptical that this is from a poor connection between the two.
I also get plenty of INFO_DST, ACKNACK between my publisher and working subscriber, so this might not be related to root cause
My faulty subscriber does not notice any publishers when I run:
ros2 topic info /topic
Have you set
ROS_DOMAIN_ID?I originally had it unset when filing this ticket.
using opensplice as my DDS implementation was working better for me (all pubs and subs noticed eachother) but once I switched to opensplice rviz started failing https://github.com/ros-planning/navig...
Hi John, I understand this is closed, or do you still have the issue ?
I never solved this issue, but I changed to opensplice so I wouldn't have to deal with it. My topic discovery weirdness went away when i switched to opensplice. and I never looked back. My setup has changed a lot since I posed this issue too, and I haven't tried to revert back to my old setup to see if the problem persists.