> Shortly after I posted the last one, I realized this problem. I cannot > > just set up a web server (or whatever) to handle the SHUTDOWN signal and > > LOGOFF. If I do, TCP/IP will just bring it back up again! I would at leas > t > need an interface to say "quit watching me, please!".
IMHO, that would be part of the stack shutdown -- as Chuckie indicated, stop the new connection listener, but also disable the auto-restart monitor at the same time -- if the stack knows it is going down, it shouldn't be resetting stuff any longer. If a server fails and the stack is in shutdown-in-progress, force/autolog won't win anyway -- nothing to connect to. The application servers would be the next targets, but if they got an EIO from socket communications, then the logical thing to do is shut down. What the individual app might do with a shutdown signal is going to be application specific; "do what makes sense", but with a common trigger to initiate that processing.