I login to the cluster and login to say N=3 nodes. Then, I laboriously open the N
panes in eg tmux or kitty, ssh onto the login node to each, then ssh onto the specific node for each of the N nodes, then start htop N times in each pane, in order to monitor resource usage on the nodes.
This is really annoying because I have log in and out a lot, if my running job crashes. Is there some way to automate this using either htop, kitty, or screen?
Alternately, is there some branch of htop that automatically monitors multiple nodes at once?
