Changing cluster always fails if engine is not manually turned off

[victorcolombo]

What's your CLI version?

0.77.0

Description & steps to reproduce

1. kurtosis cluster set docker (this triggers an Engine start locally, that spins up a container that binds to port 9710)
2. kurtosis cluster set minikube (this triggers an Engine start inside minikube)
3. kurtosis gateway(this also triggers a bind to 9710, that is already on use)

  Caused by: An error occurred running the server using the interrupt channel for stopping
  Caused by: An error occurred creating the listener on tcp/9710
  Caused by: listen tcp :9710: bind: address already in use

So the user must either kill the running engine in docker manually or run:

1. kurtosis cluster set docker
   2 kurtosis engine stop
2. kurtosis cluster set kubernetes
3. kurtosis gateway(works)

This process is laborious and error prone

Desired behavior

kurtosis cluster set X should not only start the engine on X, but also make sure that, if local, stops the local engine

What is the severity of this bug?

Painful; this is causing significant friction in my workflow.