Add method(s) to assist in graceful handling of lost RPyC connections

[loadtheaccumulator]

Add method(s) to assist in graceful handling of RPyC connections during reboots of the remote system, etc.

For example, in a reboot situation, the DeployedServer cache needs to be cleared before a new connection can be made. This can already be done via g.rpyc_close_deployed_server(), but a method to gracefully handle the failed connection, tear it down, and rebuild it would be nice.