I have this running Cloublab experiment:
https://www.cloudlab.us/status.php?uuid=eeacfb97-ca45-11ec-b318-e4434b2381fc
I have 2 nodes on there that are using pc477 hardware on the Emulab cluster, and every 30 minutes or so the nodes will become unreachable - it will display 'Connection Terminated' at the bottom of the console log, and if I try to open up the shell to get to the command line again, it will give me an unreachable error which you can see in the image attached. The only way to resolve this is for me to reboot the individual nodes. Even then, if I leave the node alone for a couple of hours, I will have to reboot it again if I want to use it.
I also want to note that when I created this experiment profile I set the hardware for those 2 nodes to 'any' sinec it doesn't really matter what hardware is used. I'm also running Linux Ubuntu 20.024 on all nodes. It is all on Emulab Cluster because I have other router nodes that need 3 network interfaces.
Also, if I have a process already running on one of the nodes, the connection will still be terminated after some time. I have not tried relaunching my entire experiemnt yet because I would need to create a disk image of everything on there and I don't want to go through that process unless it's something I really have to do to get this to work.