vr-xcon not working

[netwninja]

I'm trying to build a simple topology using 2 vEOS nodes. The nodes boot up fine no issues.

CONTAINER ID        IMAGE                        COMMAND                  CREATED             STATUS                       PORTS                                                                  NAMES
af77d840802a        vr-xcon                      "/xcon.py --p2p use0…"   5 minutes ago       Up 5 minutes                                                                                        vr-xcon-1
81ee82ae9b40        vrnetlab/vr-veos:4.22.0F     "/launch.py"             23 hours ago        Up About an hour (healthy)   22/tcp, 80/tcp, 443/tcp, 830/tcp, 5000/tcp, 10000-10099/tcp, 161/udp   manm2
4532f144d8be        vrnetlab/vr-veos:4.21.1.1F   "/launch.py"             24 hours ago        Up About an hour (healthy)   22/tcp, 80/tcp, 443/tcp, 830/tcp, 5000/tcp, 10000-10099/tcp, 161/udp   manm1

However, after building the vr-xcon container the 2 nodes are not able to to talk to each other.
I got the vr-xcon image via docker pull and tagged it as vr-xcon based on the below link reference

https://www.brianlinkletter.com/vrnetlab-emulate-networks-using-kvm-and-docker/

docker run -d --privileged --name vr-xcon-1 --link manm1 --link manm2 vr-xcon --p2p manm1/1--manm2/1 --debug

Below are the errors which we are seeing when checking docker logs

2019-06-05 04:13:12,215: xcon       WARNING  endpoint not connected, connecting to source manm1/1
2019-06-05 04:13:12,215: xcon       WARNING  connect failed [Errno 107] Transport endpoint is not connected

2019-06-05 04:13:12,215: xcon       WARNING  endpoint not connected, connecting to source manm2/1
2019-06-05 04:13:12,215: xcon       WARNING  connect failed [Errno 107] Transport endpoint is not connected

Please let me know how to fix this issue ?

Thanks
GS

[plajjan]

Sorry for late feedback (apparently putting github notification emails in a separate email folder means I don't see it).

Do you have IPv6 enabled on the docker network? I noticed that qemu doesn't listen on IPv6 so we actually have to force IPv4 name resolution. I see now I only did that for TcpBridge (that's the p2p stuff). It's sort of old commit - 0605d2e#diff-7780dd384a14cee623dd8b044ba982b0 but make sure you are on it.

You can build vr-xcon yourself - maybe that's for the best so you can make sure you get this change in there. The vr-xcon on docker hub should of course be up to date but I would need to double check that that is actually the case. You can inspect yours to see if you have this change or not.

[carlmontanari]

@netwninja I know this is old but fwiw I had exact same issue, changing the port to connect to --> 2 sorted things out. I assume because "1" is management port (though I would guess this is somewhat platform specific?):

docker run -d --name vr-xcon --link sea-eos-1 --link sea-eos-2 vr-xcon --p2p sea-eos-1/2--sea-eos-2/2

[monrad]

This is due to the mangement interface needs to be on the first PCI id so all ports have been bumped by one.
I have made a pull request to fix this #256.