how to keep workers, routers, transports processes alive after node server restart?

Let it crash, let it destroy all the data and start new.

During the crash however you should generate a new media server for users if you have one available as the server that just went down may take several minutes to return and in that time we could have one that’s alive already. :slight_smile:

Besides, if you got attacked with that type of setup, the room of users will be mad.